Cancer World 42

� Jean-Yves Blay: integrating translational and clinical research � Pioneering trialscoordinator reveals the secrets of success �Cervical cancer: closing the gap between thebest and the worst�Negotiating the principles and procedures for European biobanking

Jean-Yves Blay

Education & knowledge through people & facts

Number 42, May-June 2011

CancerWorld

42MAY-JU

NE2011

CANCER WORLD � MAY/JUNE 2011 � 1

Contents

3 EditorialAdapting services to the age of oral therapies

4 Cover StoryJean-Yves Blay: integrating translational and clinical research

13 e-Grand RoundNew approaches to treating gastro-oesophageal cancer

22 Cutting EdgeInternational biobanking regulations: the promise and the pitfalls

32 MasterpieceThe secret behind a successful clinical trial: Pinuccia Valagussa sharesthe insights gained from 40 years at the helm

40 Impact FactorHypoxia modification with radiotherapy for bladder cancerAn 18-gene signature (ColoPrint) for colon cancer prognosis

48 NewsroundCorrections & Clarifications

56 Systems & ServicesCutting unnecessary deaths from cervical cancer: collective effort aimsto narrow eleven-fold gap between worst and best in Europe

EditorKathy Redmondeditor@eso.net

Assistant EditorAnna Wagstaff

Editorial AssistantCorinne Hall

Editorial AdvisorsJacques BernierFatima CardosoFranco CavalliAlberto CostaVincent T. DeVita

Contributing WritersMarc Beishon, Simon CromptonMary GospodarowiczSusan Mayor, Peter McIntyreIain Tan, Patrick TanAnna Wagstaff

Publishing AdvisorsGillian Griffith, Fedele Gubitosi

Website LiaisonCorinne Hall

Art EditorJason Harris

ProductionHarrisDPIwww.harrisdpi.co.uk

Printed byGrafiche Porpora

Cover photographNigel Dickinson

Published byEuropean School of Oncology

Direttore responsabileAlberto Costa

Registrazione Tribunale di RomaDecreto n. 436 del 8.11.2004

All enquiries about Cancer Worldshould be made to:ESO Editorial OfficeVia del Bollo 420123 Milan, Italye-mail: magazine@eso.netTel: +39 02 8546 4522Fax: +39 02 8546 4545All correspondence should be sentto the Editor at editor@eso.net

Cancer World is published six times per year by the European School of Oncology.It is distributed at major conferences, mailed to subscribers and to Europeanopinion leaders, and is available online at www.cancerworld.org

Editorial

Theincreasing use of oral cancerdrugs is contributing to a changein the way cancer services are

organised, leading to many more patientsreceiving care in an ambulatory setting.

Cancer patients are certainly benefitingfrom this change.Many oral targeted thera-pies hold the disease in check and, if takencontinuously, can keep patients alive foryears. Reductions inhospital staysmake abigdifference to their ability to get onwith theirlives. Unlike conventional chemotherapy,side-effects associated with oral targetedtherapies are mostly mild, reversible andtend to get better over time.Avoiding needlesand the need to keep accessing central veinsis also a big plus, not to mention protectingtheir veins from the damage inflicted byvesicant chemotherapy agents.

Yet oral drugs comewith their own chal-lenges,many ofwhich are under-recognisedand poorly tackled.While the side-effects oforal targeted therapies are generally mild,they are nonetheless a burden, and all theharder to bear because of the long-termnature of the therapy. Some oral therapiesalso have complex administration schedules,which can be awkward for patients to incor-porate into their everyday life.Consequently,patients’ persistence with oral treatmentstends to drop off over time, which can havea significant impact on their outcome.

As treatment is no longer delivered inhospital, there are fewer opportunities forhealth professionals to address all theseissues and help educate patients about

� Kathy Redmond � EDITOR

adherence to treatment, managing side-effects and avoiding dangerous interactionswith other drugs or herbal therapies.

Cancer services need to adapt to makesure that patients on oral therapies do notreceive inferior care because of a lack ofinteraction with health professionals. Inmany countries, it is becoming apparent thathealth services also need to remove unhelp-ful and unjustifiable obstacles to or biasesagainst oral cancer therapies.

In some health systems, for instance,oral cancer drugs are reimbursed at a lowerrate than IV chemotherapy, with the resultthat somepatients haveno choice but to takeIV therapy, even if the overall cost of treat-ment is more expensive. The UK Parlia-ment, meanwhile, is currently debating aproposal that will make it harder for cancerpatients to access a new type ofwelfare ben-efit if they are on oral rather than IV therapy.

There appears to have been limitedhealth service planning to address themanychallenges posed by the introduction oforal therapies in cancer.Ambulatory cancerservices need to be developed to ensurethat patients’ educational and supportneeds are met, treatment-related side-effects aremanaged effectively and patientsare helped to stick with the treatment in thelong term. Reimbursement and benefit dif-ficulties also need to be addressed, so thatservices using oral therapies are not com-promised by lack of funds, and patients arenot denied beneficial oral treatmentsbecause of financial penalties.

Adapting services tothe age of oral therapies

CoverStory

4 � CANCER WORLD � MAY/JUNE 2011

Jean-Yves Blayintegrating translationaland clinical research

� Marc Beishon

Major clinical breakthroughs come from understanding molecular biology. So says EORTC

president Jean-Yves Blay, who is leading efforts to reshape cancer research so that every trial has

a translational element that can build knowledge about the mechanisms driving the disease.

Ifthe fight against cancer is mainly an incre-mental process scientifically, building care-fully on evidence, step by step, the sameapproach should apply to the people andagencies working on the problems. This

means that, in the clinic and in research, we needto keep refreshing the centres and team leaders, andgenerate a continuous streamof young investigators,to ensure that the brightest and the best are in aposition to helpmove things forward.Without this,momentum can slow down or even stop.That is the firm view of Jean-Yves Blay, the cur-

rent president of the European Organisation forthe Research and Treatment of Cancer (EORTC),whocanhimself be seenas an injectionofnewthink-ing at the head of an organisation that is at the fore-front ofmanycritical issues in oncology.Heading theEORTC is one of the most challenging jobs inEuropean cancer, as there are somany obstacles intheway of unifying research efforts around the con-tinent – not least the differences in national health-care systems and in the rules and regulations

governing research, the lack of resources and, inrecent years, the huge impact the Clinical TrialsDirective is having on academic studies and groups.And theEORTChas faced criticism–notably thatan ‘old guard’has been in place for years.“The challenges are certainly real and the criti-

cism we’ve had is fair to some extent,” says Blay,adding that the problem has been greater in someparts of the organisation than others. “We aremadeup of a number of research groups, and some arevery active, althoughothers have gone through a life-cycle where we need to bring in new blood. TheEORTC board, with director Françoise Meunierand Denis Lacombe from the headquarters, hasasked each group to identify young investigators andwe are holding meetings to help them becomeinvolved, and we have closed some groups whileothers are starting again from scratch.”TheEORTC is also lookingmuchmore towards

collaborating with other networks and agencies inintergroup studies to avoid duplicating efforts andget the most out of limited resources. “What is

CoverStory

clear though is that theEORTC is probably the onlybody with the long-term expertise to organise clin-ical research in a range of cancers across countriesin Europe,” says Blay, “and we have other groupscoming to us for this experience and not just to addpatients to their studies.”Another part of theEORTC’s strategy is to focus

on a smaller number of expert centres to carry outcomplex and demanding clinical and translationaltrial work. The Network of Core Institutions(NOCI), under discussion formany years, has nowbeen set up for this purpose. Last year agreementswere signed with core centres – there are 26 now.They will implement complex molecular trials, but

will, says Blay, also involve smaller centres, whenneeded, via EORTC’s disease-specific groups. Butsmaller centresdo lack thepatientnumbers andmul-tidisciplinary groupsneeded toparticipate in increas-ingly complex translational research studies.“Thedoor is always open, butwemust have cen-

tres capable of contributing a high level of accrualin studies, expertise in rare tumour subtypes, excel-lentmolecular biology facilities and so on– andeventhe top centres do not have all the resources andpeople on their own.There arehundreds of new, tar-geted agents in development and hundreds oftumour subtypes, withmore being uncovered eachyear. We simply do not have the resources to test

“We need to generate a continuous stream of young

investigators to help move things forward”

NIGELDICKINSON

combinations in an empirical way anymore andwemust truly integrate translational research forrational treatment development that is both effec-tive and less costly to carry out.”Blay’smain job is professor ofmedical oncology

at the University of Lyon and head of the medicaloncology department at Lyon’sCentreLéonBérard,one of France’s top dedicated cancer hospitals. Asbefits someone entrenched in the intricacies oftranslational research, he is a rare tumour special-ist, having opted to focus on sarcoma from an earlystage of his career.Thatmeanshehasbeenclosely involvedwith the

developmentof treatmentwith the standout targeteddrug Glivec (imatinib) in the treatment of gastro-intestinal stromal tumour (GIST). But as amedicaloncologist who attends virtually everymajor cancerresearch conference, he is also familiarwithmost ofthepromisingnewdrugs and theirmolecular targets,not least because some are in EORTC trials.“We are now seeing several examples each year

of new targeted agents, such as that targetingALK,with effect in both a rare tumour and a subtype ofcommon tumours such as lung cancer. Anotherexample is an agent targeting RANKL that exertsGlivec-like tumour control in a rare cancer – giantcell bone tumour– it has about 95%control, but alsopotent antitumour effect in bone metastases ofmore common tumours. We are also seeing a suc-cess story with the RAF inhibitor for melanomadeveloped by Plexxikon and now Roche. Crizo-tinib, an ALK inhibitor, is being evaluated in lym-phoma, sarcoma and other rare tumour subtypes ina NOCI trial in the EORTC. It is this kind oftranslational research –wherewe select subtypes ofpatients with different, non-standard diagnostictests – thatwe are setting our sights onnow, inwhatshould be practice-changing trials.”Most medical oncologists with strong research

interests have of course focused on biological targetselection for some time, but as Blay adds, actuallypulling together the resources to get speedy answers

CoverStory

NIGELDICKINSON

increasing number of labs and partner agencies.Sarcoma did indeed become fertile ground for

Blay, but the first major breakthrough from themedical oncology standpoint did not come untilthe identification of GIST as a distinct molecularentity, and soon afterwards the introduction of ima-tinib, which fundamentally changed treatment forthis type of sarcoma. “Then in about 2002we cameto the understanding that other sarcoma subtypesneeded different treatments based on moleculartyping and surgical classification, and that iswhat isdriving research into new agents now – under-standing themolecularbiology anddesigning rationaltreatments that target the tumour’s causation event.”AsBlay points out,molecular differentiationhas

now uncovered as many as one hundred sarcomasubtypes, and he expects at least another hundredmore. “For example, we’ve found some very raremolecular subtypes such as aEwing sarcomawheretheremay only be about 15 cases a year in Europe,while in the more common liposarcoma there arethree completely differentmolecular subtypes thatneed different treatment.”Actually treating patients is also changing, as

new agents, particularly those taken orally, arereaching clinics. “We used to do mainly in-patientclinics where people would come for two to threedays of high-dose chemotherapy. Now we havemostly outpatient clinics, as 30%of people are tak-ing only oral drugs and some IV chemotherapy canbe done in an hour or so,” says Blay. But this doesbring other pressures. It is harder for a team to keepin touch with patients and address their concerns,and in France, as in some other countries, reim-bursement for giving oral drugs ismuch lower thanfor administering IV-based therapy, which he sayscould raise funding problems in the future.France had been lagging behind some other

countries in the management of sarcoma patients,notably theUKandScandinaviancountries, saysBlay,butheadds that theFrenchsarcomagroup,whichheco-chairs,has started to followbyagreeing funding for

to the right translational research questions is stilla huge challenge. Having the right sort of platformin NOCI, together with other partners in Europeand further afield, is at least an important step for-ward he feels. His own contribution to Frenchtranslational cancer research–hehashelpedput thecountry on the international map – and his confi-dence and infectious enthusiasm no doubt playeda role in the decision to elect him to the EORTCpresidency in 2009.Blay comes from a family of doctors – including

both parents and various other relations – and hesays hewas ‘predisposed’ to go the sameway. “I didmymedical training inLyon andhadmy first place-ment atLéonBérard, and from thenon I didn’twantto do anything other than oncology – I wanted towork in a difficult areawith great promise for devel-opment and I wanted to do research.”He then spent a year at France’s top cancer cen-

tre, Gustave-Roussy in Paris, training in research,where he was fortunate to come under the wing ofthe renowned and late medical oncologist MichelClavel, who encouraged him to look around forresearch topics – and also to become involvedwiththe EORTC. “I didmy PhD on immunotherapy incancer, but also looked at tumours that were notalready crowdedwith researchers – and some thenwere almost ‘unknown lands’, including sarcoma.We could see that what we were learning aboutmolecular biology would one day translate intotreatment, but it did take some time.”Blaywas offered an assistant professor’s position

atLéonBérardandproceeded todevelopa research-basedclinical career during the1990s,working on arange of cancers, but primarily sarcoma, lymphomaand kidney, trialling high-dose chemotherapy, con-tinuing his work on immunotherapy with agentssuch as interleukin, and heading a cytokine(‘immuno-modulating’) research lab that is nowpartof a major cancer research centre. Léon Bérard, hesays, isnowsecond inclinical research inFranceafterGustave-Roussy, and inandaround thecentre are an

CoverStory

“It is harder for a team to keep in touch with patients on

oral therapies and address their concerns”

aconcertednational referralprogrammetomajorcen-tres. “This started in 2009 and is progressingwell. Itis alsomandatoryunder ournational cancer plan forall patients to have a multidisciplinary assessment,although we are still far from this target. One prob-lem ismanpower– if you aim tohave a large teamofspecialists discussingall cases, itmaybe that someoftheir time is better spent somewhere else. But wehave shown in a small local study in Lyon thatpatientswhodo reachamultidisciplinaryboardmayhave abetter progression-free survival – about20%.”About 10% of French sarcoma patients are now intrials, he adds, although in theUK, one of theEuro-pean leaders, it stands at about 18%.Now, he adds, a

much larger study isunderway comparing themanagement of sarco-mas in the Rhône-Alpesregionwith national rec-ommendations andwithother regions in bothFrance and Italy, again todemonstrate the impacton survival. This work isled by a certain memberof his group at LéonBérard, namely IsabelleRay-Coquard, a medicaloncologistwho also hap-pens tobemarried toBlay, inwhatmustbeoneof theclosest personal/professional partnerships in Euro-pean oncology.That study iswork thatwaspart ofConticanet, a

‘networkof excellence’for connective tissue tumoursfundedby theEuropeanUnion’s sixth frameworkpro-gramme (FP6), andwhichhasnowended.Blaywasthe coordinator of the project, which aimed toimprove the molecular characterisation, manage-ment and treatment ofwhatmany think is a raredis-ease group, but which does affect up to 10,000people a year inEurope.Asusualwith theseprojects,whichcompriseanumberofworkpackages,Blay says

whatwasactually achievedwasdifferent to some ini-tial proposedoutcomes: “But I suspect something isnot goingwell if it’s not changingorevolving,”he says.Two particular changes he mentions are includingmolecular subtypes in cataloguing the epidemiologyof sarcomas in Europe, and integrating academicresearch on surgical and radiotherapy treatment,and also imaging. Blay helped set up aEurope-widepatient group as part of the project, the SarcomaPatients EuroNetAssociation (SPAEN).These framework programmeprojects have the

aim of leaving a sustainable legacy, and one Blay isparticularly pleased about is a virtual tumour bank,which so far has 10,000 paraffin and fresh-frozen

sarcoma samples (seealso Cutting Edge,p22). “This is accessi-ble by anyone in theworld, with agree-ment for researchbased on each con-tributor, and shouldbe a good model forrare cancers – anyonecan contribute pro-vided theymeet qual-ity guidelines suchas central pathologyreview, and we willsee if it will be used

to identify new treatment targets.”Even without Conticanet, Blay says that the

world of sarcoma researchers has been particularlyclose and there is a gooddeal of cooperation onwhodoeswhatwork. Researchers from some countries,China, for example, have recently joined clinicaltrials for the first time, andhe saysmultinational col-laborations are increasing rapidly.And clearly, as thecomplexity of translational research increases, theissue of how centres cooperate to answer the rightquestion in any cancer type will be crucial.“I have done a lot of translational research, and

much of it has been what we call descriptive, or

“This tumour bank is accessible by anyone in the

world and should be a good model for rare cancers”

CoverStory

Partners. With a group of patient advocates at the firstConticanet Patient Advisory Group workshop, Paris, 2006

SIMONBACONNIER

temozolomidewas found to benefit mainly patientswith an inactivated repair gene calledMGMT.The most well-known translational study co-

ordinatedby theEORTC is, of course,MINDACT,large-scale research using gene expression cap-tured on microarrays to improve the selection ofbreast cancer patients who can be sparedchemotherapy. “I know there are strong opinionsabout it, and I accept that some seeMINDACTasa complex academic exercise, but it is important notonly because of the question but also for demon-strating that such research is feasible at this level,”says Blay. “If it shows improvement in patient sur-vival, that will be a real proof of the clinical value ofthe gene expression concept, and if not, thenwewillgo in another direction.”He is still sceptical though of the applicability of

gene expression in routine practice, although hedoes not doubt its value. “It offers an integrated viewof genes being expressed in the cancer cell, but itmay be hard to apply ultimately in clinics outside ofmajor centres. What may be reproducible are the

more harshly, cosmetic, wherewemaynot understand the mechanisms, andof coursewemust carry on supportingthis work. But now we need muchmore to build what I call integratedtranslational research, which is ouraim at the EORTC and was set intrain bymy predecessors as president,Martine Piccart and Lex Eggermont.”Taking his ownwork to explain the

progression in translational research,Blay says one of his most cited earlierpapers found that patients respondingto immunotherapy with IL-2 (inter-leukin-2) did less well if they wereoverexpressing a factor called IL-6dur-ing treatment. “This is probably trueand has been reproduced by others,but it did not go much further. Butanother study looked at a serum test toidentify patientswhowould not benefit from treat-ment and we were quite successful in also corre-latingVEGF level with a lack of response. Thiswasconfirmed later by other studies, not just those onimmunotherapy, and is an important phenomenonthat may have contributed to the development ofVEGF inhibitors in kidney cancer, in particular incombination with interferon.“Then we had another example, the EORTC

phase II trial of imatinib inGIST,where in one armwe showed therewas90%control inGISTandnonein non-GIST sarcoma. That was proof that select-ing patients on the basis of a molecular alterationmade sense. And finally an example of truly inte-grated translational researchwas finding thatGISTswith a certain uncommon mutation (PDGFRA)were not going to respond to imatinib, so we cannow identify these patients and drive them toanother protocol where there is response.”A recent example that has been amajor transla-

tional research success for theEORTCis in glioblas-toma (a high-grade brain tumour), where the drug

CoverStory

“We need much more to build what

I call integrated translational research”

NIGELDICKINSON

Research Council to take a lead on projects, forexample,while theNOCI grouping, he says,will bea “truly efficient network instigating new trials thathave molecular alteration as an inclusion crite-rion.”NOCI is, though, currently dominatedby cen-tres in northern Europe and especially France, theUK and the Netherlands, and Blay acknowledgesthat there is still toomuch fragmentation in researchnetworks aroundEurope. TheEORTC itself has abudget of less than€20million – a far cry from itsmain international partner, the National CancerInstitute in the US.Nevertheless, the EORTC, which is funded

mainly by charity, is probably the most successful,long-standing cancer research network on the con-tinent – andBlay intends it to stay thatway. “It is crit-ical we involve more young investigators so thatEORTC research groups do not losemomentum–and it’s a great way to develop a career as they willbecome better known to colleagues in their owncountries.” The quality of younger researcherswhodo get involved at European level is good, but it is aself-selected group– “It is a challenge for oncologiststo keep up with the latest in molecular biology.”Another area of fragmentation that concerns

Blay is what he terms the ‘double culture’ of basicand clinical scientistswhen it comes to translationalresearch. Unusually for a clinician, perhaps, hewould like to see more invested on the basic side,while maintaining clinical levels. “Major clinicalbreakthroughs come from understanding molecu-lar biology, but we have to bring the two groupstogethermore.” Biologists need to understand that,while they have complete control of experiments intheir labs, they have to adapt to reality in the clinic,where say collection of specimens is subject towhat surgeons and pathologists can provide, andethical and legal constraints. Networking meet-ings for clinical and basic researchers are now keyevents in Lyon, he says.Like many senior oncologists, Blay’s own

research interests have expanded rapidly during

“Blay talks of seeing encouraging signs that

industry is investing in European centres”

CoverStory

structural alterations ofDNA– there are easier toolsto test things like amplifications from paraffinembedded tissue.”So one direction for research is the smaller trial

onmore focused populations of definedmolecularsubtypes,where the value of large-scale randomisedtrials canbe less useful, saysBlay, pointing to the tenor so key targets now in play, such asRAFandALK.“If the population is not homogeneous enough, ofcourse we will be doing large randomised con-trolled trials, butwemay not need them for certainagents with a targeted population when an out-standing clinical benefit is observed, imatinib beinga prime example.”A thread common to both types of trial is

increasing organisational complexity, given theneed to link a wide pool of researchers working indifferent countries and continents, often across dif-ferent cooperative networks, to accrue both expert-ise and patients, especially for rare subtypes. AsBlay points out, there are two main types of clini-cal trial – drug development, where there is usuallyindustry input, and therapeutic strategies, whichreally need multidisciplinary approaches, and forwhich funding is hard to find for independent aca-demic research.Althoughmuch early-phase drug development

is done in theUS, Blay talks of seeing encouragingsigns that industry is investing inEuropean centres,thanks to the quality of the researchers, “and pos-siblywe are better placed for the focus on accruingpatients for subtypes – a well understood molecu-lar pathway is a goodway to get tomarket now.”Andwhile major obstacles still stand in the way of aca-demic research, theEORTChas foundways tomit-igate the worst effects of the Clinical TrialsDirective, which reduced the number of EORTCtrials to amerehandfulwhen it first came into force.The Directive is now up for revision.Blay is keen to stress that the EORTC will be

networking more widely on research, and will behappy for agencies such as the UK’s Medical

A close partnership. Blay’s wife,Isabelle Ray-Coquard, chairs thegynaecological clinical research groupof INCA, the French cancerresearch body, but she alsoshares Blay’s interest insarcoma and GIST

‘subvert’ the immune system to their ownadvantage.He has long been a human dynamo, constantly

on themove–colleagues speak of phenomenal driveand unlimited energy, despite him seemingly nevereating lunch, which is very unusual in France. Heis said to be a superb diplomat in the ‘oncopolitics’of the European research community, taking onmany organisational taskswith charmand tact (andnever saying ‘no’). Combinedwith his leading posi-tion in clinical and translational research, thismakes him a key player across several fronts.Close international colleagues include Paolo

Casali inMilan andAllan vanOosterom inBelgium(both sarcoma experts) and Jaap Verweij, a topmedical oncologist in Rotterdam. In France, men-tors have includedClavel, and two now retired keycancer centre directors, Thierry Philip at LéonBérard and Thomas Turz at Gustave-Roussy.As if being in perpetual motion on the cancer

frontwas not enough, Blay has a family of four chil-dren and finds time to also listen to a huge musiccollection and sometimes go snowboarding.Will cancer research have a downhill run to

success?Blay is certainly intent onmoving things onas fast as possible. “I want to see true personalisedtreatment in selected groups in routine use in fiveyears’ time – that way we will really improve

survival – and I want to see theEORTC and others doingmore of the high-qualitytranslational research weneed to do this.”

his later years and he is probably one of Europe’smost published lead and co-authors, with morethan300papers, principally on sarcomas,GISTandimmunotherapy, but also on public health issuessuch as breast and prostate cancer screening pro-grammes.Aquick glance at hisCV though canmissthe fact that in 1999hemoved from the cancer cen-tre to head the oncology unit at the nearbyEdouardHerriot university hospital,where in anine-year spellhe experienced being just another specialist com-peting for attention.“When I arrived, somewelcomedmebut others

said things like, ‘I’m an organ specialist doingresearch on cancer – I don’t need an oncologist’ –and it was very challenging to convince people towork together as we built up an in-patient depart-ment for chemotherapy there. I had the highermission though ofmerging its cancer activities andthat of other hospitals in Lyon with Léon Bérard,and we have been quite successful, particularly inrare tumours. We now bring together physiciansfromdifferent sites into a sarcoma board sowe canallocate patients according to the best surgical spe-cialists, for example.”However, in countries wherereimbursement is based on activity, andmoney candisappear when referrals to other hospitals aremade, there canbemajor obstacles to setting up thistype of treatment networking, notes Blay.Since 2009, Blay has been back at Léon

Bérard, and has the flexibility to devote atleast a day a week to his EORTC work,along with clinical and research activities.He is studying very rare sarcomas,while hisimmunotherapy research has movedon to reconstituting the immunesystem rather than attacking thetumour itself. “It’s completely dif-ferent from the immunotherapywork of the 1990s,” he says.The work has included show-ing that breast cancer cells

CoverStory

PIERREHEUDEL

e-GrandRound

New approaches to treatinggastro-oesophageal cancer

Robust evidence on the value of neoadjuvant chemotherapy, together with more effective

imaging modalities, are opening up new options for diagnosis, staging, treatment and patient

selection in gastro-oesophageal cancer. Andrés Cervantes looks at the implications for the

management of this challenging disease.

T he challenges of managinggastro-oesophageal cancer areillustrated by the case of a 50-

year-old man with locally advancedoesophageal cancer. His main present-ing symptom was dysphagia when eat-ing solid foods. He had weight loss, nopain in the thorax or abdomen, no dys-phonia or cough, no dyspepsia or gastro-oesophageal reflux. His performancestatus was considered to be 1.

Looking at his risk factors, thepatient was a heavy smoker, he had sig-nificant – though moderate – alcoholconsumption, and was obese, with abody mass index of 32.5 kg/m2.

Physical examination revealed nolymph nodes in the patient’s neck nor inthe supraclavicular area; there werealso no thoracic alterations, nohepatomegaly or ascitis, and no signs ofpleural effusion. All these physicalexamination data indicate that thepatient had localised disease. Havingdysphagia only when eating solid foodsindicated that the invasion of theoesophaguswas not completed becausethe patient could swallow liquids with-out any difficulty.

The European School of Oncology pres-ents weekly e-grandrounds which offerparticipants the opportunity to discussa range of cutting-edge issues, fromcontroversial areas and the latestscientific developments to challengingclinical cases, with leading Europeanexperts in the field. One of these isselected for publication in each issue ofCancer World.In this issue, Andrés Cervantes of theHospital Clinico Universitario, Valencia,Spain, provides an update on the chal-lenges of treating gastro-oesophagealcancer. He highlights the new treatmentoptions and techniques for predictingtumour response that are changing thetreatment of patients with this cancer.Florian Lordick of the Braunschweig

Clinic, Germany, poses the wide range ofquestions sent in by participants duringthe e-grandround live presentation, whichis summarised here by Susan Mayor.

The recorded version of this and other e-grandrounds is available at www.e-eso.net

DIAGNOSTIC TESTSAn oesophagoscopy indicated an ulcer-ated neoplastic lesion in the lower thirdof the oesophagus at 38 cm from theteeth, situated over an ectopic areaof gastric mucosa, 3 cm long, withoutinvolving the gastro-oesophageal junc-tion. This oesophagitis was related togastro-oesophageal reflux. This descrip-tion is very clearly aBarrett’s oesophagusarea transforming into a lower thirdoesophageal cancer.

The stomach and the duodenumwere also explored because the tumourcould have passed through the stenoticlesion, but they did not showany change. During the proce-dure, a biopsy was performedand this showedpoorly differen-tiated and infiltrating diffuseadenocarcinoma of the oesoph-agus over Barrett’s oesophagus.

The oesophagogramwas oneof the most common diagnostictechniques performed someyears ago, and we still performthis type of imaging. The image(see figure) shows some alter-ation of themucosa of the lowerthird of the oesophagus, withirregular areas indicating amalig-nant tumour of the oesophagus.

A CT scan of the thorax,abdomen and pelvis showed nodistantmetastases; the lungs andliverwere clear, and the only twofindings were related to local regions.First, in the oesophagus, there was athickness of thewhole oesophageal wallat the lower third of the oesophagus.Thiswaswithout any anatomical relationto the trachea or left bronchus. Therewas also enlargement of lymphnodes inthe para-oesophageal area and in theupper mediastinum.

After doing the oesophagoscopy, theoesophagogram and a complete bodyCT scan, the clinical staging is consid-ered cT3cN1cM0 stage 3a (where ‘c’

indicates clinical assessment), becausethe tumour has completely thickenedthe whole wall of the oesophagus. It isN1 because the enlarged lymph nodesare observed in the CT scan and cM0because, clinically, there is no evidenceof metastatic disease. We consideredthis patient to be at clinical stage 3a.

To summarise this, I would like toconsiderwhatwas theclassical approachto oesophageal cancer. First, the patientunderwent surgical resection. After sur-gical resection, pathology assessmenthelped in the estimation of the risks.Postoperative treatment was based on

the classical TNM stage. Postoperativechemotherapy was of either doubtful orno value, andpostoperative chemoradia-tionwas recommendedby someexperts.

Question: What do you consider thestandard procedures in staging oesophagealcancer today?Which examinations can berecommended as standard?Answer: I think that oesophagoscopy isthemain procedure for diagnosis, althoughthis does not help to stage the patient.Wewould use a CT scan and then a PET

e-GrandRound

scan, especially for a tumour located in thelower third of the oesophagus and at thegastro-oesophageal junction. I think CTandPET scanning are the critical tests.Wedo not use ultrasonographic endoscopyvery often, except for patients with verysmall tumours – patients with T1 andT2involving only the mucosa or muscularlayer – but they are uncommon. In apatient at such an early stage, perhapsendoscopic ultrasonography could be ofbenefit in diagnosis.Lordick: We use endoscopic ultrasono-graphy to define theT stage.Would you sayit is also goodwhenyoucombine endoscopy

and PET-CT scans? Would thisbe your standard approach?Answer: That is right.Question: In your case presenta-tion, the patient hadnomajor dys-phagia. If the patient presentswithdysphagia, do you insert stents or doyou dilate the oesophagus beforeyou startwith any other treatment?Answer:Assessing the nutritionalstatus of a patient with oesoph-ageal cancer is very important.This patient was presenting withno weight loss, indicating thatalthough dysphagia was themainsymptom, it was not presentingany problems. In patients pre-senting with major dysphagia wewould try to do staging proceduresvery quickly because, when thedisease is localised, improvement is

common when you start chemotherapy.However, whenwe see a patient with dis-seminated disease, we try to start byimplanting a stent first, because treat-ment is unlikely to be very successful andwe prefer to focus on improving theirnutritional status.Question:What about the use of protonpump inhibitors with prokinetic drugs inorder to prevent carcinoma of the oesoph-agus in patients with gastro-oesophagealreflux disease?Are you aware of data thatshow that oesophageal cancer can be

DIAGNOSTIC OESOPHAGOGRAM

Oesophagrams should be used strictly for diagnosis and not forstaging. The irregular dark area shows the malignancySource: Courtesy of Andrés Cervantes

standardised uptakevalue (SUV) of 8.9 g/ml.Another metabolic areawas observed at theupper mediastinum(SUV max: 3.4 g/ml),associated with thelymph glands (seebelow). This confirmedthat this patient hadlocoregional diseaseinvolving the oesopha-gus and also locore-gional lymph nodes.

DISTRIBUTIONOF GASTRO-OESOPHAGEALTUMOURSSquamouscell carcinomaof the oesophagus ismainly located in the upper two-thirds oftheoesophagus, andaccounts for approx-imately 18% of cases. Adenocarcinomaof the lower oesophagus affects 26% ofpatients, while adenocarcinoma of thegastro-oesophageal junction accounts fora further 26%. Gastro-oesophagealtumoursat these twosites are increasing inincidence and are common in developed

countries – the US andEurope – where gastriccancer is not so common.Traditionally, gastric can-cerwasmore common inMediterraneancountries,but its incidence is nowdecreasing,while tumourslocated around the junc-tionand the lower sectionof the stomach areincreasing. In this clinicalcase, smoking, alcoholand obesity are all relatedto the reflux disease,which couldbe the causeof the tumour.

The Siewert classifi-cation of gastro-oesoph-

prevented by the long-term use of protonpump inhibitors?Answer: I am not aware of any data thatshowthat thosepumpsmaypreventpatientsfrom developing oesophageal cancer.

CURRENT APPROACHESTO LOCALISED GASTRO-OESOPHAGEAL CANCERAmultidisciplinary approach should beadopted in the care of patients withlocalised gastro-oesophageal cancer.Clinical staging should include PETscanning for tumours located in thelower oesophagus or in the gastro-oesophageal junction.

The patient in our case study had aPET scan performed with labelledfludeoxyglucose (18F), which confirmedthe absence of distant metastases. Thisis a very important point because PETscanning is more sensitive even thancurrent helicalCTscans indetecting dis-semination of the tumour.Approximately18%–20% of patients without metasta-sis in the conventional workup are diag-nosed as stage IV with a PET-CT.

The 18FdG PET-CT scan showed ahypermetabolic area at the lower thirdof the oesophagus, with a maximum

ageal junction adenocarcinomas is a sur-gical one: type I is when the tumour islocated in the distal oesophagus (36% ofpatients); type II is true junctionaldisease(27%); and type III iswhen thebulkof thetumour is belowor at the subcardial area(37%) (Br J Surg 85:1457).

PREOPERATIVE CHEMOTHERAPYSeveral studies have helped us in under-standing that preoperative chemotherapyshouldbeconsidered the standardof care.

Ameta-analysis of randomised clin-ical trials of preoperative chemotherapyfor oesophageal cancer by Gebski et al.(Lancet Oncol 8:226) showed clear,although limited, benefits (see p 16).Many of the studies were underpow-ered, with a very limited number ofpatients, and are now relatively old.

The MRC trial OE02 is importantbecause it included800patients,makingit the largest trial published so far in gas-tro-oesophageal cancer. It is also themostrecent trial, published in the Lancet in2002 (vol 350, p1727). The studyincludedpatientswith all types of resect-able oesophageal or cardia carcinoma,

e-GrandRound

18FdG PET-CT SCAN FOR STAGING

Areas of very high uptake of the 18FdG show the tumour in the lowerthird of the oesophagus and reveal the involvement of lymph nodesin the mediastinal areaSource: Courtesy of Andrés Cervantes

DISTRIBUTION OF GASTRO-OESOPHAGEAL TUMOURS

These relative incidence data were adapted from the USSurveillance, Epidemiology, and End Results programme(SEER), and are likely to be different in other parts of the world

including squamous carcino-mas, adenocarcinomas andundifferentiated carcinomas(n=802). The patients wererandomised to surgery aloneor to the experimental arm oftwo courses of cisplatin(80mg/m2 onday1)and5-flu-orouracil (1000 mg/m2 con-tinuous infusion days 1–4),given in a very conventionalway and repeated after 21days, with surgery performedimmediately after the secondcourse. It is important tonotethat two-thirdsofpatientshadadenocarcinoma histology(66%) and one-third hadsquamous histology (31%).

Results (see below)showed that median survivalwas improved by 3.5 months(13.3 vs 16.8 months), andoverall survival after two yearswas improved by 9% (34 vs 43 months).Most types of histology benefited fromthe experimental approach (Lancet350:1727). Long-term results – after 10years of follow-up – have recently beenpublished, indicating that the benefits ofpreoperative chemotherapy continuedover the longer term(JCO27:5062–5067).

Thenext trial Iwould like todiscuss is

theMAGICtrial.Althoughmanypeopleconsider this trial as a gastric cancer trial,a group of patients with oesophagus andjunction cancer were included. Patientswere randomised tosurgeryonly, or topre-operative chemotherapy with the classi-cal British ECF combination offluorouracil, cisplatin and epirubicin.Patientswere given three courses of pre-

operative chemotherapy fol-lowedby three furthercoursesafter surgery. The quality ofthe trial is well established,and it has been published inthe New England Journal ofMedicine (vol 355, p11).

Final results indicate thatmedian survivalwas improvedby four months (24 vs 20months) and two-year sur-vival increasedby9% (50%vs41%). The key finding was a13% increase in five-year sur-vival (36%vs 23%).Bothpro-gression-free survival andoverall survival were signifi-cantly improved.This benefitapplied to all patients, includ-ing those with stomachtumours, junction tumoursand oesophagus tumours.

A further study of peri-operative chemotherapy for

localised gastro-oesophageal cancer, theFrench trial FNLCC 94012-FFCD9703, has not yet been published in full(Boige et al, Abstract 4510, ASCO2007). It randomised 224 patients tosurgery alone or to three courses of plat-inum and fluorouracil (5FU800mg/m2

on days 1–5), and did not include epiru-bicin, making it different to the British

e-GrandRound

META-ANALYSIS OF NEOADJUVANT CHEMO TRIALS

Preoperative chemotherapy was shown to offer a slight advantage in thismeta-analysis, which included studies dating back to 1982Source: Reprinted from Gebski et al (2007) Lancet Oncology 8:226-234,

with permission from Elsevier

UK MRC OE02 TRIAL

Ten-year results showed a sustained improvementin both disease-free and overall survival withpreoperative chemotherapyCS – neoadjuvant chemotherapy plus surgery

arm, S – surgery alone

Source: WH Allum et al (2008) JCO 27:5062–

American Society of Clinical Oncology.

ASSESSING TUMOURRESPONSE TO TREATMENTThe classical way of assessing responseto chemotherapy is anatomic imaging.After several weeks, or even months, oftherapy we expect to see a reduction intumour size. This could be seen as avery late effect, six to eight weeks afterstarting treatment. This is whatwe do atthemoment, butwe have to improve onthis. We have to make the most of theopportunities we have now to includebiochemical ormolecular imaging. PETscanning is useful because it can detectan effect of therapy in reducing cell pro-liferation or in stimulating cell death.This effect can be seen as early as one totwo weeks after treatment, so it canguide treatment decision making. Sev-eral groups have published data on this,and more work is ongoing.

AstudybyOtt andcolleagues, lookingat tailoring treatmentsbasedonmetabolicresponse (Ott et al JCO 24:4692)included 65 patients with locallyadvancedadenocarcinomasof thegastro-oesophageal union who received preop-erative cisplatin-based chemotherapy.They were assessed before the start oftreatment andonday 14.Results showaclear difference in PET response, based

trial. Another difference is that two-thirds of the patients had gastro-oesophageal junction tumours, and onlyone-third had gastric tumours.

Results showed that perioperativechemotherapy improved progression-free survival and overall survival.Mediansurvival increased by nine months andthree-year survival by 10%, while five-year survival increased by 14%.

These three trials of perioperativechemotherapy for localised gastro-oesophageal cancer are all leading us inthe samedirection.They showthe reduc-tion in the risk of death at five years was25% (in the Cunningham trial; NEJM355:11) to 31% (in the French trial;Boige,ASCO2007), indicating that peri-operative chemotherapy may improvesurvival when given to patients withlocalised and resectable oesophageal orgastro-oesophageal cancer.

Ameta-analysis of randomised clini-cal trials of preoperative chemoradio-therapy for oesophageal cancer (LancetOncol8:226) indicatespotential benefits,but also flags up some issues. There is atrend to higher mortality in patientstreatedwithacombinationofchemother-apy and radiation after surgery.Althoughthis is nowacommonlyusedapproach inthe US, my personal opinion, which isshared with others, is that in Europechemotherapy without radiation couldbe the standard care for patients withlower-third, junctionandgastric cancers,before surgical resection.

Question: Inwhich patients do you con-sider using neoadjuvant chemotherapy?Answer: In all resectable patients exceptthose clinically staged as cT1.Question: Would you use neoadjuvantchemotherapy in T3 tumours and alsowhen you see lymph nodes involved?Answer: Yes, for any node-positivetumours, includingT3 andT4,wewoulduse preoperative chemotherapy.Question:Do you use, or is it possible to

use, capecitabine and oxaliplatin as part ofthe neoadjuvant chemotherapy?Answer: Several trials performed inadvanced disease have shown that oxali-platin can be used safely and with goodefficacy, substituting for cisplatin.Capecitabine could also substitute for flu-orouracil. Going back to the clinical casewewere reviewing, we recommended pre-operative chemotherapy, and a combina-tion of oxaliplatin and capecitabine wasstarted. After the first course, the patienthad complete resolution of dysphagia anda reassessment with PET-CT was per-formed after twoweeks to assess themeta-bolic response.Question:Doyou think that chemoradiationis the preferred approach over chemotherapyalone, in neoadjuvant indications?Answer:Theonlyplace inwhichchemora-diation could be better than chemotherapyis in patients with squamous tumours ofthe two upper-thirds of the oesophagus.Definitive chemoradiation can cureabout 25% of these patients. Apart fromthis group of patients, as far as I know,there are no good randomised trials com-paring chemotherapy with chemoradia-tion. The data we have from themeta-analysis compared chemoradiationwith surgery alone. There are also sometoxicity concerns. For thisreason, Iprefer to startwithchemotherapy alone inpatientswith tumours thatare resectable, andnot usechemoradiation.There areseveral studies showingthat chemoradiation iseffective, and can result inan evenhigher proportionof patients with patholog-ical complete remissionthan using chemotherapyalone. However, I thinkthis strategy shouldbe con-sidered experimental, andshould be explored infuture phase III trials.

e-GrandRound

ASSESSING TUMOUR RESPONSE

New imaging modalities offer the potential to identify respondersmuch sooner than relying on tumour shrinkage

patient. The main data from the casestudy’s pathology report showed thepatient had a moderately differentiatedadenocarcinomaof the loweroesophagus.The tumour was infiltrated to the peri-oesophageal fat, so we considered thispatient to have a ypT3 tumour (where‘yp’ indicates pathological assessmentafter preoperative therapy). Perineuralinvasion was present, although none ofthe ninemediastinal nodes and none ofthe 13 perigastric nodes were involved.So this patient was ypT3 ypN0/22 M0out of the 22 resected lymph nodes. Inthe omentum, no metastatic depositswere detected in the peritoneum.Over-all, this patient was considered as aypT3 ypN0/22 M0 patient.

ona reductionof at least 35% in themax-imum standard uptake value. Thesepatientsdobetter than thosenot showinga significant responseonPET. I think thisis an important finding, although it isbased only on retrospective studies.

In the patient in our case study, earlyreassessment with a PET scan per-formed two weeks after starting treat-ment showed a complete response,withno uptake of 18FdG. Knowing that thispatient was so chemosensitive, wedecided to go on and give four morecourses of the combination ofcapecitabine and oxaliplatin.

Results from theMUNICONphaseII trial looking at the value of PET toassess earlymetabolic response andguidetreatment of adenocarcinomaof the gas-tro-oesophageal junction confirmed thefinding that patients with an early PETresponse that leads to several courses ofchemotherapydomuchbetter than thosenot showing a PET response, who godirectly to surgery (LancetOncol8:797–805). These data confirm theprognosticvalue of PET scanning in the assess-ment of response to treatment.

SURGICAL RESECTIONThe patient in our casestudy underwent thoraco-abdominal surgery with atotal oesophagectomy. Anoesophago-gastrostomy wasperformed using Akiyama’stechnique, with extensivemediastinal and perigastriclymphadenectomy. DuringreconstructionusingAkiyama’stechnique, an omentectomywas also performed.

PATHOLOGYASSESSMENT ANDESTIMATION OF RISKIt is very important to assesspathology to be able to esti-mate the risk for a particular

POSTOPERATIVE TREATMENTMany trials have been designed withpre and postoperative treatment. How-ever, the postsurgery state of patients,including their nutritional status andthe presence of surgically related com-plications, means that almost half ofour patients cannot go through thisfinal part of therapy. We decidedagainst continuing chemotherapy tothe patient in our case study onaccount of the changes to his nutri-tional status.

The patient is being followed upwith clinical visits every three to fourmonths for two years. He has no dys-phagia, his trachea is working welland he can eat properly without any

difficulties in swallowing.No postoperative chemo-therapy was given due to thepatient’s poor nutritionaladaptation and slow recov-ery after surgery. He lostaround 20 kg in weight aftersurgery, although no steat-orrhea or other signs of mal-nutrition were observed.There is no standard of carefor follow-up, and thisshould be based on clinicalsigns and symptomsreported by the patient. Werepeated a CT scan everysix months for two years,and this patient has notshown any evidence ofmetastatic disease or signsof local relapse so far.

e-GrandRound

PET AS A GUIDE TO TREATMENT

The MUNICON trial showed that metabolic response assessed early on aPET scan can help doctors identify which patients will benefit frompreoperative chemotherapySource: Courtesy of Andrés Cervantes

RECOMMENDED APPROACH FOR LOCALISED OESOPHAGEAL CANCER

� Clinical assessment and staging� Multidisciplinary team discussion� Preoperative treatment in all

patients with clinical stage II and IIIdisease

� Surgical resection after chemotherapy� Pathology assessment and the

estimation of risk� Postoperative chemotherapy?� Participation in trials

CONCLUSIONSIn conclusion, the patient in our casestudy was diagnosed with a locallyadvanced lower third oesophageal adeno-carcinoma.Clinically, therewasaT3node-positive tumourwithabsenceofmetastaticdisease. Iwould like to stress that amulti-disciplinary discussion is essential for all

cases. This case is a good example of thebenefits ofmultimodality treatment. Thesurgical approachallowedanR0resection,which is essential to offer the patient thepossibility of long-term survival. Aresectable oesophageal cancer should betreatedwithpreoperativechemotherapy inamultidisciplinary team approach.

More, andbetter designed, clinical trialsare needed to refine the optimalapproach. Iwould encourage everyone toconsider entering your patients intoappropriatemulticentre trials to providemore information on the optimalapproach formanaging oesophageal can-cer for the future.

e-GrandRound

Florian Lordick (FL) of the Braunschweig Clinic, Germany, hosteda question and answer session with Andrés Cervantes.

Q: When do you consider postoperativeradiotherapy?A: That is a very difficult question. If theresection isR2orR1, and there is evidenceof microscopic or gross residual diseaseafter surgery, then radiation could be con-sidered as a palliative therapy. But weshouldbalance thebenefits of the radiationagainst the potential of inducing someother toxicities. I would be careful inselecting only those patientswith residualdisease after therapy for postoperativeradiation. In general, it is not my stan-dard of care.Q:What type of chemotherapy should beused? Is it necessary to include epirubicin inthe treatment regimen?A: In the three trials indicating positiveeffects, two gave positive results for thebenefit of chemotherapy without usingepirubicin. Most data on epirubicin arefrom theBritish group and arewell estab-lished findings, but we have other trialsindicating that, even in the absence ofepirubicin, there are benefits withchemotherapy. So the use of epirubicin isnot a must in my opinion.Q: Given that early PET response is pre-dictive, when the patient responds tochemotherapy can you go with definitivechemoradiation and avoid surgery?A: This approach has not beenwell stud-ied in patientswith lower-third adenocar-

cinoma. In general, it is considered thatsurgery may be better than radiation asdefinitive treatment in these patients. Iwould recommend definitive chemoradi-ation only for those patients with cancerslocated in the upper two-thirds of theoesophagus. Even in the presence of agood response on PET scan, I would rec-ommend surgery as standard of care.FL: I agree that in patients with adeno-carcinoma of the distal oesophagus thereis not yet a clear role for definitivechemoradiation. The standard approachtoday is surgery for patients presentingwith resectable adenocarcinoma of thedistal oesophagus and gastro-oesophagealjunction.However, I think the hypothesisis justified and studies should be con-ducted to see whether there is a role fornon-surgical treatment in very good PETresponders.Q: Do you also use radiation withoutchemotherapy as preoperative treatment inoesophageal cancer?A:No, I think radiation alone should notbe used as preoperative treatment inoesophageal cancer. There are data andrecommendations in the current ESMOguidelines showing there is no indica-tion for radiation alone, apart from pal-liative treatment in some advanced andunresectable disease, but not as neoad-juvant therapy.

Q:Does neoad-juvant chemo-therapy in oeso-phageal cancerimprove survival?A: I showed ameta-analysis and,moredefinitively, threetrials, that indicate short-term and long-term improved survival in patients withgastro-oesophageal cancer receiving pre-operative chemotherapy.These three trialsgive evidence at level 1 that preoperativechemotherapy does improve survival inpatients with oesophageal cancer.FL:Weshould consider that this questioncame from India, where there may bemorepatientswithoesophageal squamouscell cancers than in Europe. Maybe thequestions asked in India are not answeredby the trials presented andwe needmorestudies in that part of the world.A: The squamous situation is different.Only the MRC trial included patientswith squamous cancer (31% of the total).However, there was a very good trial pre-sented by Kelsen et al. in 1988 (NEJM339:1979–1984) indicating that preoper-ative chemotherapy in patientswith squa-mous disease is not so beneficial.Q: Do you agree that there are differencesbetween the optimal management ofpatients presenting with squamous cellcancer and those presenting with adeno-

e-GrandRound

carcinoma of the oesophagus?A:Yes, I agree.Forpatientspresentingwithsquamous cell cancer, the aetiology andbiology are different, and the type ofpatient is different – it is more related tosmoking. So, palliative care for thesepatients would be definitive chemoradia-tion and I would resort to surgery only ifthey do not respond or if the patient’stumour is still resectable. These types ofdisease are complicated.Q:Do you think there is enough evidenceto use oxaliplatin instead of cisplatin in thepreoperative setting?A: The data we have on oxaliplatin arevery limited, so I do not think that thiscould be considered a standard of care.However, data from many trials show itsadvantages when oxaliplatin is substi-tuted for cisplatin in the treatment ofmetastatic disease. Low-dose oxaliplatinis noweasily accessible, and not as expen-

sive as it was some years ago. Overall, Iwould not consider there is level I evi-dence to substitute oxaliplatin in locore-gional disease, but for practical reasons Ithink we could use it.Q:After surgical treatment, do you alwaysadminister postoperative chemotherapy evenin cases of yPT1 N0 and no more risk fac-tors, after complete resection?A: I do not have the definitive answer tothis question, but if a patient has a com-plete resection and is chemosensitive Ido not see any reason not to use it, solong as the patient has adapted well aftersurgery and is keeping their weight stablewithout treatment. I try to treat all patients,if their pathology reports are good, withthree more courses of postoperativechemotherapy. However, only half ofpatients are in a good enough condition toreceive this type of treatment.Q:Howmany lymph nodes are required to

consider that surgery was successful, andwhat do you do if there are only a few lymphnodes in the specimen. What is yourapproach?A: I am not aware of any guidelines indi-cating the number of lymph nodes wehave tohave in the resected specimen.Butfor the junctionwe shouldhavemore than14 in order to get the right staging. Whenthere is residual disease after chemother-apy and after surgery, the use of postoper-ative chemotherapy is indicatedwhateverthe number of lymph nodes present.Q: Is it possible to insert a stent preopera-tively if necessary?A: We never do that. This would only beconsidered in a patient with rapid wors-ening of dysphagia. I would try a nasogas-tric tube or just change therapy if a patientis sensitive to other drugs. But, in general,most patients improve after just a fewdaysof chemotherapy.

Learning can be fun

e-ESO sessionsJoin us every Thursday and the first Tuesday of each

month from 18:15 to 19:00 CEST (Central European Summer Time)

For further information please visit www.e-eso.net

is also specified as an essential activity formembers of the elite Organisation ofEuropean Cancer Institutes (OECI).AEuropean network for data-linked

frozen cancer specimens TuBaFrost,was set up in Rotterdam in 2003, and isnow in the hands of theOECI. Special-ist networks, such as Conticanet forconnective tissue cancers, are buildingup their own banks to identify molecu-lar subtypes. A number of countries,including Sweden and the UK, havealso embarked on projects to developpopulation-based biobanks, whichshould enable researchers to study sam-ples from cancer patients (and others)that were taken while they were stilldeemed healthy, to look for biomarkersof early detection or risk.A considerable number of data-

International biobanking regulations:the promise and the pitfalls

� Anna Wagstaff

Moving towards personalisedcancer therapy is about find-ing answers to questions like:

Is this colon cancer aggressive or fairlyindolent?Which type of chemotherapywill this breast cancer respond to?Doesthis person need to take preventativemeasures to guard against a raised riskof cancer?Finding those answers involves look-

ing at samples taken fromaggressive andindolent (or responsive andnon-respon-sive) tumours and identifying abiological‘marker’ or ‘biomarker’ that appears todifferentiate the two. Further samples,fromnewly diagnosed tumours, are thenneeded to test, in a prospective study,whether such ‘candidate’ biomarkersreally canpredict thebehaviourof thedis-ease and can thus be relied on to help

guide theclinician towards the right ther-apy for their patient.The raw materials for all this work

are large quantities of quality-controlledand well-catalogued biological speci-mens linked to information about theperson from whom they came, theirhealth status and the trajectory of thedisease. A shortage of these raw mate-rials will slow down progress in improv-ing cancer treatments.Responding to this need, the

research community has been steadilybuilding up ‘biobanks’ as repositories ofdata-linked human biological samples.Systematic banking of samples forresearch is becoming increasingly com-mon at major cancer centres and uni-versity hospitals and is nowmandatory inmany. Biobanking samples for research

CuttingEdge

Banked samples of human tissue, blood and serum linked to the patient’s clinical data are the

raw materials of modern cancer research. A single infrastructure for Europe’s rapidly evolving

biobanks is urgently needed. But finding agreement on the ethics, language and operating

standards is proving quite a challenge.

ILLUSTRATION:FREDVANDEELEN,WWW.ORGANISART.CO.UK

CuttingEdge

that research groups like the EuropeanOrganisation for Research and Treat-ment of Cancer (EORTC) run along-side clinical studies, on the principle of‘no tissue, no trial’. This principle, putforward by ESMO’s José Baselga –now Professor ofMedicine at Harvard– holds that it is a waste of resources toorganise a trial that answers the ques-tion: ‘Is a better/no worse than b?’, if it

fails to gather biological samples thatcould answer the personalised ther-apy question: ‘For which patients is abetter/no worse than b?’The blossoming of these biobanks

is very good news, and yet their abilityto serve the research communityremains limited by their fragmenta-tion. Each has grown up with itsown set of norms, principles, quality

linked biological samples have alsobeen collected as part of specific stud-ies. The MINDACT study, for exam-ple, collected and gene-profiled frozensamples as an integral part of the pro-tocol, which aims to see how accu-rately theMammaprint gene signaturecan predict who will benefit from adju-vant chemotherapy. Other collectionscome from ‘correlative’ translationalresearch studies – ancillary protocols

The blossoming of these biobanks is very good news,

yet their value remains limited by their fragmentation

the consent formwould require approvalfrom an ethical review board. In somecases, researchers may even be obligedto re-contact the donor to get a newconsent. This disparity in conditionsattached to the use of samples presentsa potential obstaclewhere, for instance,a research centre from a country withstricter rules wishes to join an interna-tional biobanking system where tissuescould be transferred for use in a countrywith more liberal rules.

DEBATES OVER PERSONAL DATABiological material, important though itis, makes up only half the researchequation. The other half is data thatdescribe the donor (age, gender etc)and disease (eg stage IIIb non-small-celllung cancer), as well as data that allowresearchers to draw conclusions aboutthe significance of biological differ-ences with regard to disease progres-sion, response to treatment or perhapsadverse effects.Though there is a European Direc-

tive on data protection, the laws andpractices governing the storage and useof personal data for research varywidelyacrossEurope – as has been highlightedby the battles that have had to be foughtin some countries just to get the go-ahead to set up a basic cancer registrythat could link a person’s cancer diag-nosis to their cause of death.As with the consent issue, there is a

broad consensus around a general prin-ciple: namely that patients have a right tokeep their medical details confidential,and that biobanks may therefore onlystore anonymised data. There is lessconsensus, however, onwhat thismeans.

NO CONSENSUS ON CONSENTThe principle that an individual shouldhave some say over whether and howtheir biologicalmaterial is used iswidelyrecognised within research fields andthewider community, but is interpretedin a variety of ways. As a result, onebreast cancer tissue samplemaybe avail-able for use in anywell-founded researchstudy, whereas a similar samplebiobanked elsewhere may have tightrestrictions on its use. Consent mayhave been given purely for use in thestudy for which it was originally col-lected, or for additional specified studies,or perhaps with the proviso that use inany studies beyond those specified on

CuttingEdge

standards and IT and legal frameworks,often in response to specific needs.The question is how tomove to a moreharmonised system that would enableany authorised researcher with a studyproposal to instigate a single searchrequest across all relevant biobanks,together with agreed rules, guidelinesand principles governing the collec-tion, storage, transfer and use of thesedata-linked samples. As if this werenot challenge enough, it has to beachieved within the relevant nationaland European rules and regulations,including those on data protection andon the rights of the individuals fromwhom the samples are taken.

The principle that an individual should have a say over

the use of their biological material is widely recognised

The raw material of cancer research. Large numbers of fresh-frozen tissue samples like these, linkedto clinical data, hold the key to learning about the biological differences between cancers, andunderstanding what these differences mean for the way a given cancer behaves and thus the treatmentstrategies that will be most effective

Some countries interpret it in thetightest possible manner – each donoris assigned a code that is used to iden-tify their data and their biologicalsamples, and the link between thedonor and the code is then destroyedto preclude the possibility that some-one could access the banked data toillicitly look at an individual’s privatemedical records.This is fairly disastrous for the pur-

poses of research, because it meansresearchers cannot go back to the treat-ing oncologists to get updates on howa patient’s disease progressed or how itresponded to various treatments. Theycannot even go back to ask for addi-tional information – on side-effects forexample – that might have been avail-able at the time of anonymisation, butwas not deemed relevant.Many countries, however, do

accept ‘two-way’ coded data as‘anonymised’, if there are sufficientsafeguards. With two-way coding thepatient’s oncologist, or a third party,keeps hold of the code book, thusretaining the possibility for researchersto request further information aboutthe donor. How easy that process isdepends on how stringently the systemis safeguarded. In the chain of tissueand data, double coding (coding addi-tional to the one done at the sourceinstitution) may be required and per-mission to decodemay involve complexand possibly bureaucratic procedures.Such discrepancies between coun-

tries on the level of personal data pro-tection presents another potentialreason why some countries may notwant their citizen’s data to be interna-tionally available through a biobank.

achieve this. Some, however, sound anote of caution. Evert-BenVanVeen, anexperienced medical lawyer at Med-Lawconsult in the Netherlands, pointsout that the Clinical Trials Directivewas an exercise in harmonising rulesand regulations across Europe – andlook how that turned out.Van Veen carried out extensive

research on the various rules and reg-ulations governing procedures for con-sent and data protection across Europewhen the ErasmusMedical Centre in

CuttingEdge

A CASE FOR HARMONISATION?GettingEU-wide agreement on a singleset of rules governing consent and dataprotection – or indeed other social/eth-ical issues such as duties to publish theresults of any research done using thesesamples, andwho should own the intel-lectual property rights – may seem theobvious solution. The Biobanking andBiomolecularResourcesResearch Infra-structure (BBMRI), an initiative to buildan infrastructure for Europe’s biobanks,is currently working on proposals to

“I realised that if you want to make such a model, you

end up with the regulations of the strictest countries”

A PROLIFERATION OF GUIDELINES

Numerous guidelines and recommendations have been published regarding the ethical andsocial issues in biobanking.� In 2003 UNESCO issued the International Declaration on Human Genetic Data. Among

many other things, it set down the principle that “prior, free, informed and express con-sent, without inducement by financial or other personal gain, should be obtained for thecollection of human genetic data,” and that “Human genetic data, human proteomic dataand biological samples linked to an identifiable person should not be disclosed ormadeaccessible to third parties, in particular, employers, insurance companies, educationalinstitutions and the family.”

� In 2006 the Council of Europe issued its recommendation Rec(2006)4 of the Commit-tee of Ministers to Member States on Research on Biological Materials of Human Ori-gin, which laid down more stringent conditions for consent, particularly for unspecifiedfuture research, than those currently in force in many EU member states.

� In 2009 theOECD issued its Guidelines for HumanBiobanks andGenetic ResearchData-bases, which appears to have been written principally with healthy volunteers in mind.The level of detail required in the consent forms seems inappropriate for cancerpatients, from whom samples are taken within routine diagnostic and treatment pro-cedures, when they will have many other things on their minds.

� In 2010 the Organisation of European Cancer Institutes (OECI) published its ethical andlegal recommendations, From the Biobank to the Research Biorepository, with an empha-sis on building public trust and support. They recommend that biorepositories take theform of charitable trusts or other ‘neutral’ bodies with amission to act in the public inter-est, and that they develop policies for the development of patents from research carriedout on samples…“with the aimof protecting the public interest to enjoy new technologiesfor health at reasonable costs.”

Rotterdamwas in the process of settingup TuBaFrost in 2003. “The idea wasthat we would come up with a har-monised model for exchanging tissue:what would be accepted in one coun-try would also be accepted in othercountries.” It wasn’t long before a flawemerged in the strategy. “I realised thatif you want to make such a model,then you end up with the strictest reg-ulations of the strictest countries,because what is accepted for theless strict will not be accepted bythe stricter. That would be very harmfulfor research, especially for thoseresearchers who have started theirbiobanks with lighter regulations.”The Council of Europe has, in fact,

already agreed a ‘Recommendation toMember States on Research on Biolog-ical Materials of Human Origin’,Rec(2006)4, which Van Veen claimscontains provisions considerably stricterthan those operating inmanyEuropeancountries. The recommendation thatany consent forms be “as specific as pos-siblewith regard to any foreseen researchuses”, for instance, appears to rule outasking patients to consent for their sam-ples to be used in unspecified futureresearch projects – an option widelyseen as essential in cancer, where therate of change in knowledge and tech-niquesmake it hard to foresee all possi-ble research projects.Any researcher wishing to use sam-

ples for a researchproject “notwithin thescope of prior consent”, say the recom-mendations, should make “reasonableefforts” to contact the person in order toobtain consent to the (new) proposeduse (with further conditions to be ful-

filled if that fails).Given that the samplesare coded, that many years may haveelapsed since the original consent, andthat it is cancer patients we are talkingabout here, this processwould probablybe not only complex and time-consum-ing, butmay even causedistress to a can-cer patient or family.Were those recommendations given

legal status, this would hinder efforts ofthe countries most actively promotingbiobanking, some ofwhich, theUKandSweden for example, do give patients anoption of allowing residual tissue to beused in future for unspecified research.Some even have an ‘opt out’ system,whereby patients are not asked for con-sent, but instead are given an opportu-nity to indicate that their tissue shouldnot be used for research purposes. Thissystem operates in Belgium and Den-mark, with the Netherlands expectedsoon to follow suit.

BIOBANKING IS A SOCIAL ACTIVITYVan Veen is highly critical of what hesees as the ‘paternalistic’ and ‘conflict-based’ approach taken by most docu-mentsonregulatingbiobanking, includingtheOECD’s2009Guidelines forHumanBiobanks and Genetic Research Data-bases (see box, p25). They are, in hisview, driven by the instincts of civil ser-vants to regulate everything, they lackany democratic basis, and above all theyfail to recognise that citizens don’t justwant protection, they alsowant, andwillbenefit from, progress in medicalresearch. He says that patients are oftenvery keen to be partners in research,pointingout that inBelgiumand theUK,twocountrieswhereproposed legislation

sparkedwidepublicdebate, the laws thatwere finally passed were much moreresearch-friendly than theoriginal drafts.He believes Europe’s approach to

biobanking should be rootedmuchmorefirmly in the solidarity-based values thatunderpin the continent’s healthcare sys-tems, such that “the healthy contributepart of their income to the sick, theyounger to the older, etc.”“What iswrongwith expecting some-

one to contribute to observationalresearch when it does not affect theirpersonal lifeplan and other patients willprofit in the longer run?” asks VanVeen.This approach is apparent in the

OECI recommendations for the opera-tion of research bio-repositories, whichsays that, “to ensure compliancewith thewishes of donors,” every biobank shoulduse the samples in thepublic interest. “Inno way can samples be considered, orbecome, ownedbyprivate for-profit enti-ties.” It also recommends that everybiobank should “disclose its rules, activi-ties and results to the scientific commu-nity and the general public…topromotea culture of solidarity and consent todonations.”Rather thanasking theEUtopass yet

more rules and regulations, Van Veensuggests it should insteadadopt a generalframework for good researchgovernancebasedonkeyprinciples consistentwith asolidarity-based health system. Theseshouldencompass issuesof transparency,accountability and thenon-profit basis ofbiobanking, the right to opt out as amin-imum, aswell as “how thegeneral resultsof researchwill bedisseminated, ‘conflictof interests’ policies, how the issues ofintellectual property rights aredealtwith,

CuttingEdge

“Europe’s approach to biobanking should be

rooted much more firmly in solidarity-based values”

riore Sanità in Rome, and he runs thebiobankat theNationalCancerCentreatBari, southern Italy,which systematicallycollects andbanks data andbiomaterialsfrom every patient. This system, heargues, is farmore valuable than collect-ing samples only from specific trials,because it assembles samples from theentire cancerpopulation, rather thanonlyfrom patients selected by age, or diseasestage or other criteria.Paradiso recently completed anexer-

cise that has introduced a single ‘lan-guage’ and software system throughoutthenetwork.He isnow looking towork incollaborationwithothernational andpan-European networks to establish condi-tions formoving towards a similar level ofharmonisationacrossEuropeandbeyond.The term ‘language’ covers many

issues that the clinical research commu-nityhaswrestledwith for years. Differenthospitals may use different thresholdsfor judging a tumour tobeER-positive, ordifferent tests for establishing theHER2status. Evenmenopausal statusmay notbe defined in the sameway.Paradiso mentions also the stage of

howtheconfidentialityofpersonaldataofdonors is maintained, etc.”Above all, heargues, this should not become an extrabureaucratic layer.This leaves the question of how

countries with more strict conditionscould agree to participate in an interna-tional biobank where samples may beused in countries with less stringentrequirements. The answer, Van Veensuggests, is to use the ‘coordinating prin-ciple’ adopted by theTuBaFrost project,which states that, wherever the researchsamples are actually used, theymust behandled in accordance with the regula-tions of the country where the tissuewas taken from thepatient and originallystored. This would permit unhinderedexchange of biological samples withoutputting unwanted barriers in the way ofbiomedical research.Avoiding a new setof EU regulations would also make iteasier for individual countries to reviewtheir own rules through their own dem-ocratic procedures.

A COMMON LANGUAGEIf it is best to leave ethical and socialdetails to countries, there are otheraspects of biobanking for which thereverse is true.One big challengewill beestablishing a common ‘language’for cat-aloguing samples to ensure that whatappears in the search results correspondsto what the researcher is looking for.This issue has been preoccupying

many cancer research leaders, includingAngelo Paradiso, who is himself deeplyinvolved in the effort to find biomarkersfor early diagnosis and for predictingresponse to treatments.Paradiso is theco-ordinator for all of Italy’s cancer centrebiobanks, togetherwith the IstitutoSupe-

disease: “If I want to compare the char-acteristics ofmysamplewithothers com-ing from other tumour banks, I have toclassify the tumour stage, histologicaldiagnosis and cytohistological grade inthe sameway.”Then there is the question of how far

you go in defining a tumour for the pur-poseofa searchablecatalogue?Evena rel-atively rare cancer type such as sarcomaisnowknown toconsist ofmore thanonehundredbiologicallydistinctdiseases (seecover story), and it is becoming increas-ingly apparent that different biologiesbehave very differently.

STANDARDS FORHANDLING AND STORINGAgreement on basic quality standardsandquality control of thecollection, stor-age and transfer of samples is anotheressential element, so researchers can beconfident that, nomatterwhere the sam-ples originated, their studies will not beconfounded by poor-grade samples.Sophisticated techniques for gene

profiling or proteomic andmetabolomicstudies can be highly sensitive to small

CuttingEdge

If samples are to be catalogued and exchanged across

borders, there has to be an agreed classification system

Paraffin-embedded or fresh-frozen?Fresh-frozen tissue is needed for screening approaches like gene expression profil-ing or proteomics,which are used to search for biomarkers or to learn about themech-anisms of disease. These techniques are highly sensitive and strict criteria areneeded about collection and storage.FFPE (formalin-fixed paraffin-embedded) blocks used in standard pathology are farless sensitive to discrepancies in theway they are collected and stored. They are use-ful because large collections already exist from trials going back decades.The initial research behind theOncotypeDXmulti-gene assay that gives risk scoresfor certain breast cancers was done using gene profiling in frozen tissue, but muchof the validation work was done retrospectively on FFPE tissue blocks.

differences in samples. The minimumstandards recommendedby IARC/WHOin 2007 are widely accepted as a goodstartingpoint.However, theydonotcoverissues suchaswhatdrugs (not just cancerdrugs) the patient may have been on atthe time the sample was taken, or thetechniques used for the sampling. Fur-thermore, these types of study all requirefresh-frozen tissue, with the time from‘harvesting’ to snap freezing being one ofthequality parameters, andquestions arenowbeing askedaboutwhether the timeshould be measured from the point ofexcision or from the point of clampingduring theoperation, asdepriving the tis-sue of oxygen induces rapid changes.This in turn raises the question ofhowmuch you can ask of operatingteams, for whom annotating sam-ples is not their top priority.

TOWARDS INTERNATIONALBIOBANKINGThe goal of reaching agreementbetween Europe’s biobanks mightseemhopeless, given thedisparities inpractice and the difficult balancing actbetween cataloguing ‘essential’ informa-tionwithout demanding toomuchof thepathologists anddata processors – not tomention the disparities in the softwareused to input that information. But Par-adiso is confident it can be done. “Whenyou talk about biobanks, you should talkabout networks,” he says. While har-monising every biobank for every diseasein everyEuropean countrymight seemabig ask, if it is done network by network,working from the national level up, thetask becomes a lotmoremanageable.The framework for these networks

has already been developed at a Euro-pean level by the BBMRI in the form of‘hubs’ that group different types ofbiobank at national and then Europeanlevel. “Take Italy, you now have the hubfor population-based biobanks, the hubfor cardiovascular, for cancer, for geneticdiseases and so on. This is the first level.The second is at the international level,where all national hubs take part, con-nected in a common platform, in whichall kinds of communication is possible –within a hub and also between hubs.”As has happened in Italy, each of the

hubs at national (first) levelwill of neces-sity work towards a common ‘language’

and set of minimum data and qualitystandards, fromthebottomup,with all ofthem hopefully following the interna-tional discussions and trying to movetowards a harmonised system that couldfunction internationally. “Discussion andagreement has to be reached, first at onelevel and then the next,” says Paradiso.To aid this process, he adds, there are

already tools that make it possible toaccept data fromanyof the software com-monly used in hospitals and biobanks.

Software, however, is no substitute foragreement between themajor networks.To this end, the Bari Cancer Centrehosted a meeting for biobank networkslast November, attended by representa-tives frommore than30 organisations inEurope,Asia andAfrica. Itwas organisedby theOECI together with ESO, underthe auspices of the EORTC andendorsed by ISBER (the InternationalSociety for Biological and Environmen-tal Repositories).“Themainaim,” saidParadiso, “was to

share experiences from all these groups,and to discuss the main possibilities forbiobanking from a clinical perspective.What do biobanking organisations needin termsofminimumstandards andmin-imum data requirements?”These can be easy questions to

answer, comments JacquelineHall, whorepresented EORTC at that meeting,but only if you know what study youwant to carry out. “If you know, forinstance, that youare going tobecol-lecting a serum sample to be usedfor proteomics profiling, you knowalreadywhat the goal is and you can

already have in mind key variables youmight want to collect about how thatsamplewas taken fromthepatient orhowit was processed, because there areknown factors that can influence theproteomics profile.“In the case of unspecified future

use it becomesmore tricky, because youhave to findwhat is practical for the localpathologists and hospital staff to pro-vide, and what is practical for managingthe data here at EORTC HQ, and bal-ance that against the needs of theresearch effort. As soon as you start

CuttingEdge

EORTC is developing templates for what it considers

to be the key information for different sample types

the efforts being put in nowwill result inexisting biobanknetworks gelling into aninternational system. As new biobanksjoin, this will transform the accessresearchers have to data-linked samplesand significantly speed progress inunderstanding cancers and how todetect, diagnose and treat them.For this to happen, guidelines and

regulationsmust not only harmonise thebiobanks. They must also inspire confi-dence in thepublic, inpatients and in theclinicians and pathologists who are atfront line of collecting samples, that thewhole enterprise is based on the princi-ple of solidarity, where the gains fromthese voluntary donations are dissemi-nated and used for the public good.

With groups like this sharing experiencesanddiscussing commonpositions on lan-guage, key associateddata, andminimumquality requirements, the foundations arebeing laid for a biobank that can operateon a truly pan-European level. But Par-adisohas sethis sights on reaching further.

Present at theBarimeetingwererepresentatives from Egypt,Tunisia, Israel and Jordan, all

keen to develop biobanking intheir countries.A follow-upmeet-inghas been schedulednext year in

Romania,with a focus onpro-moting a biobanking cul-ture in central and easternEurope, and beyond.The best scenario is that

collectingmore data it ismorework andmore cost. So we try to find a trade off.”Collecting samples is now a perma-

nent concern for the EORTC and a pri-ority for their research studies. In Jan-uary theorganisation released anupdatedpolicy on Human Biological MaterialCollection, Storage and Use, that coverssample collection for both specifiedand unspecified use (see Policy pageat eortc.be). This policyincludes makingavailable for ‘sec-ondary use’ thedata-linkedsamples itholdsin an inde-pendentlyrun biobankfacility inMilan,as well as at variousinstitutions that par-ticipate in EORTCstudies. Hall says it isalso developing templatesfor what it considers to be thekey information for differentsample types. “We discuss thesevariableswith thepeople involvedin the studies, the pathologistswho collect the samples and ourown pathobiology group, in thecontext of international collabo-rations. We also look to infor-mation in the public domainaboutwhich variablespeoplefind important for differentactivities, and then we havean internal discussion to findout which will be the key vari-ables for the different sam-ple types in the contextof that study.”

CuttingEdge

Half amillion people – around 1 in 50 aged between 40 and 69 – respondedto the invitation to ‘join the BiobankUKproject’ (www.ukbiobank.ac.uk). They

attended 21 centres across England, Scotland and Wales togive blood, urine, saliva, and a variety of clinical measure-ments and filled out questionnaires on their lifestyle andmedical history. Trust in the public service values of the

NHS, which supports the Biobank, played an important rolein motivating people to take part. The central messagefocused on the opportunity to take part in an exciting

research project, and the idea that samples may beused for an unspecified purpose in future was pre-sented as an opportunity rather than a threat: “In10 or 20 years’ time the things that we will beable to analyse in the samples may well be

things that scientists have not yet thoughtabout. The next generation of scientists,whomight still be in primary school today,will actually use new tests and new

methodologies to be able to unlock new secrets interms of how we prevent diseases.”

Guidelines and regulations must not only harmonise the

biobanks, they must also inspire public confidence

BIOSOCIAL CITIZENS

The secret behinda successful clinical trialPinuccia Valagussa shares the insights gained from 40 years at the helm

� Simon Crompton

Goodclinical trials areproposedbyclinicians, have thepotential for real patient benefit and increase

knowledge about thedisease. So saysPinuccia Valagussa, who forgets tomention another secret

of success: having someone like herself in charge, who works closely with clinicians, keeps tight

control over the quality of data, and is dedicated to helpingmore andmore centres join trials.

Thefirst thing thatPinucciaValagussa says tomewhen Imeether in the receptionof the IstitutoNazionaleTumori inMilan is that shedoesn’t

want todo this interview.This is a little disconcerting,though she says it in a very friendly, polite manner.Then, thank goodness, she leadsme down the corri-dors toher office, explaining that of course shewill doit, so that she can get over importantmessages aboutclinical trials andcurrentbarriers to good research. It’sjust that the more she’s thought about the interview,themore she’s feared it.The problem is that Valagussa, awomanwhohas

been at the centre of somekey trials in the recent his-tory of cancer research, hates talking about herself.She agrees to interviews thinking it flattering, butthen has second thoughts because, she says, she is avery private person.Her dislike of the limelight isn’t affectation.Dur-

ingmost of our interview,Valagussa,who isdirector ofthe Operations Office for Clinical Trials at the

MichelangeloFoundation inMilan, speaksopenlyandanimatedly–discussing thequalities of good trials, thebureaucracy that stifles significant research, andsomeof theexciting studies shehasbeen involvedwithover40years. She is all expressivehands, facial contortionsandacombinationofboth thatmakes Italiansuniquelyable to express “that’s how it goes”, “what canyoudo?”and “I told you so” all in one go.Yet when I venture into her background, motiva-

tions and influences, all that stops. “I really don’tknowwhat to tell you,” becomes a regular reply.Her demeanour is perhaps not unexpected given

that she has had a central role in 350 papers on sys-temic adjuvant therapy for early breast cancer, treat-ment of malignant lymphomas and methodology inclinical trials, yet hernamehas rarelybeen first in listsof authors.Valagussamaybea lynchpin to someof themajor advances inclinical oncologyover fourdecades,and shemay have received several awards (includingan ItalianWomanof theYearAward in1997andaCity

Masterpiece

DEDICATED TO CLINICAL TRIALSValagussa explains tomehow theprinciple aimof theMichelangelo Foundation is to design and conductclinical studies and translational research. Free ofcharge and independently, it assists clinical oncologyinvestigators fromtheearliest planning stages. “Wegothrough all the administrative burden, ask other sitesto join the investigation, discuss the objectives andscheme of the study, go to the regulatory authoritiesandethics committees, collect all thedata, assess thequality of thedata, planandconduct theanalysis, andprepare for presentation and publication.”The rigour the foundation applies to planning

studies assures a quality of research that is farmore likely to have an impact on clinical practiceand patient care than studies that are poorlydesigned or never get off the ground because of

of Monza scientific merit award in 2005), but sheworks in the background.She has run the Operations Office for Clinical

Trials at the IstitutoNazionaleTumori inMilan since1973, seeing its clinical trials office develop in 1999into theFondazioneMichelangelo, anon-profit organ-isation devoted to advancing research in cancer. In2007 she became a director of the foundation. Theofficewhere shehasworkedsince thestart is in theoldpart of IstitutoNazionale Tumori. Thewalls are cov-ered with prints of impressionist paintings – thechoiceof renownedcancerdoctorGianniBonadonna,who founded theMichelangeloFoundation.Despitehaving had a disabling brain haemorrhage in 1995,Bonadonna is still theheart and soul of theoperation:his book-smothered office is next to Valagussa’s, andhe greetsmewarmlywith a left-handed hand-shake.

Masterpiece

ELIGIOPAONI

time-consuming administra-tive procedures. Valagussa’soffice ensures that nothingcoordinated by them com-promises its high standards.“Agoodclinical trial is first

of all one that is proposed byclinicians, because they haveideas. The hypothesis, ifproved, must show a benefitthat is clinically important andimportant to the patient. Forexample, is it important to starta very large studywith the aimof finding a difference of nomore than 3% between treat-ments a and b? You mayimprove the rate of survivalbut a new treatmentmay alsohave risks. There’s a dangerthat such studies are likecom-paring Coca Cola with PepsiCola: is the goal really to benefit the patient?“Agoodclinical trial improvesknowledgeof thedis-

ease, and it is importantnowadays thatwhendesigninga clinical study you have to keep inmind that youwillneed to correlate it with a translational study. So youneed to talk to yourpatients andexplain the importanceof themdonating samples for future research.”

AN IMPRESSIVE TRACK RECORDThe research that Valagussa andher teamhave beeninvolved in over thedecades demonstrates the poten-tial impact of well-planned clinical research. In theearly 1970s, with Bonadonna, she coordinated thelandmark trial showing that adjuvant CMF(cyclophosphamide,methotrexate and fluorouracil)provided significant survival benefits for womenwith operable breast cancer – a finding that hasbeen confirmed in follow-up studies over 30 years. “Itwas quite a departure. We demonstrated to sur-geons howpatients could be curedwith chemother-apy, so it began to change mentalities, and was the

beginning of the multidisciplinary approach.”Another landmark trial occurred in theearly1970s,

when Bonadonna designed a new combinationchemotherapy forHodgkin’s diseaseknownasABVD(adriamycin,bleomycin, vinblastineanddacarbazine).With Bonadonna, she coordinated the trial that in1974 showed the superiority of ABVD comparedwith thestandardMOPP(mecloretamine, vincristine,procarbazine, prednisone) chemotherapy. ABVD istoday still considered the gold standard for conven-tional chemotherapy inHodgkin’s disease.In the late 1980s, she coordinated trials under

Bonadonnaandwith thesupportofUmbertoVeronesi,whichchallenged theclassic indicationofmastectomyfor breast tumours of three centimetres or more,demonstrating that primary chemotherapy beforesurgery reduced tumour size, and that conservativesurgery could be an effective and safe alternative toradical surgery. “Inwhatwas, andstill is, a surgical cen-tre, wewere able to say: ‘Please, now, we can all helpour patients preserve their body integrity by starting

Masterpiece

Outstanding young investigator1985. Valagussa’s contribution tosetting up some of the first big

breast cancer trials was recognisedwith this prize awarded by the

Italian Health Minister

“A good clinical trial is first of all one that is

proposed by clinicians, because they have ideas”

with chemotherapy followed by surgery.’ It took awhile to convince people, but wemanaged it.”There’s much more to come. She is currently

planning a global trial of a new type of adjuvant ther-apy, involving groups fromItaly,Britain andAustralia,but the details as yet have to be kept under wraps.There have been massive changes in the scope of

trials into cancer drugs over the four decades thatValagussa has been at the Istituto Nazionale Tumori.She first came in 1969, a Red Cross volunteer nursebasedatMonzaHospital attendingacancercourse thatthe Institutewasholding.Because shehadstudied lan-guages at high school, spokeEnglish andhadattendedcourses in statisticswhile inMonza, shewas asked byVeronesi to join the Institute as a scientific secretary.“Nobody really toldmewhat theywanted fromme

untilmy first day in the job, whenProfessor VeronesitoldmeIwouldbe involved in trials. Ididn’t knowwhatthis meant. I never associated the word trial withmedicine before.”Soon shewas thrown into compiling information

forVeronesi’s trial onbreast cancer surgery, and typingupBonadonna’sprotocols forchemotherapy trials.Theclinical trials operations office, officially set up toconcentrate onmedical oncology in 1972, was origi-nally a small affair. It coordinated only single-centrestudies for the Institute itself andhad just three staff.Nowthereare12staff, coordinating studies in35cen-tres around theworld.

GOING MULTICENTREIts growth can be traced to 1993, when Bonadonnadecided to respond to requests frommedical oncolo-gistswhomhehadtrained,andwerenowworkingelse-where in Italy: they wanted to participate in some oftheclinical studieshewas running. “Itwaswith somereluctance initially,” says Valagussa, “because you areused toworkingwithinyourowngroup,and it’snot thateasy to change.But itwas an important step, becauseit meant we would be able to cooperate togetheraccording to certain rules, and itwould allowpatientsfromother regions of Italy to have good experimental

treatments, based on sound clinical reasoning, with-out coming toMilan.”So around 15medical oncologists from northern

Italy got together for an exploratory meeting at theMichelangelo Hotel near Milan train station; theydecided to stay in touch, and called themselves theMichelangelo group. They did indeed start multi-centre trials, coordinated from the trials office of theIstituto Nazionale Tumori, and years later, when thework of the office became formalised into the newfoundation,Bonadonnadecided to continuewith theMichelangelo name.Themove to internationalmulticentre trials came

in themid 1990s, when Bonadonna was designing anew randomised trial to test classical adjuvantchemotherapy against neoadjuvant primary chemo-therapy before surgery in cases ofmoderate- to high-risk breast cancer. One of the drug companiesproviding funding askedwhether itwouldbepossibletoconduct it as an international trial. Soaprotocolwas

“We can all help patients preserve their body integrity

by starting with chemotherapy followed by surgery”

Masterpiece

A phenomenal partnership. The collaboration between Valagussaand Gianni Bonadonna, one of medical oncology’s great leaders,has not just improved survival and quality of life for countlesscancer patients but helped set the standards for clinical research

arrangedandplansweremade.And thenBonadonnahad a brain haemorrhage.“Wehad a big discussion in Paris with the investi-

gators and the drug company, and we had to askwhether we could continue this adventure withoutDr Bonadonna. Finally, Professor Luca Gianni, thendirector ofmedical oncology at the IstitutoNazionaleTumori, accepted thechallenge.So in1996,westartedthe internationalisationofour foundation.Andoncewedid it, we knewwe could do it again and again.”International trials suddenly presentedValagussa

andher colleagueswithnewchallenges for organisingconsistent protocols. Different countries had verydifferentperceptionsofwhat ‘best conventional treat-ment’ was. There were different technological levels– somecentres participating in trials in the late1990sdidnot evenhave routine access to the internet.Drugcompanies funding the trials had tobeasked formoremoney to help less well-resourced sites participate.Achieving quality data in these large trials is time

consuming. “It’s costly, but not just financially. It’s notalways easy to get investigators to send the right kindofdata at the right time– theyhave their job todo in theirclinic, after all.And you need to convince them of theimportanceof following rigorously all the safety proce-dures in your protocol. And if something doesn’t lookright to you in the data, you need to call the investiga-tors anddiscuss itwith themandprovide advice.Whatqualifies our team is the clinical quality of the data.While themainpriority of a drug companymight be toensure that all the right boxes in the study have beenfilled, and thismight bedone at the endof a study, ouremphasis right from the start is to check the quality ofthedata. It’s not so important that information ismiss-ing. It’s important that what you have is good.”

THE BURDEN OF BUREAUCRACYBut the biggest challenges have always beenposed by bureaucracy. It’s a problem that afflictsresearchers in every country, but Valagussa believesinternational trials are battling against almost impos-sible odds to get off the ground. Even a specialisttrials office such as her own struggles with the

convolutions of red tape that drain time andmoney.If all goes smoothly in planning for a large study, itwill take at least four years to complete enrolmentand many more years to follow-up. In that time,other findings and developmentsmay havemade astudy’s original objectives obsolete. Valagussa saysthe situation is sometimes “nightmarish” for organ-isations like her own attempting independentresearch driven by the needs of patients.“The regulatory authorities are all different in

every country involved.Youhave to get your protocolcleared with them, and then present to the ethicscommittee, and then you have to select the partici-pating sites.Nowadays, things are gettingworse. Forexample, for a non-profit organisation like ourselves,conducting a non-profit study, it is not clear underEuropean ruleswhether you, as the sponsor, have topay for the drugs used in the study, evenwhen theyhave been registered for the indicateduse. It seemsto be different in different countries.“According to European regulations, sponsors

have toprovide a fee to the regulatory committee anda fee to theethics committee.Theyoftenhave topayfor all the drugs, and sometimes for extra patientexaminations. If this continues, what is the futurepossibility of academics and institutes like our ownconducting studies? They are just too expensive.”So it is inevitable that funding fromcommercial

sources has to be accepted for many studies. Vala-gussa’s office tries to help researchers find inde-pendent sources of funding, but these rarely coverthe full cost of a study. Sincemuch of the researchis to establish new indications for drugs that havealready been approved, drug companies are askedfor support too. But Valagussa emphasises thatthere can be no drug company intervention in stud-ies’ design or objectives.What could be done to make quality, inde-

pendent multicentre research easier to accom-plish? Valagussa shakes her headwearily. “I haven’tany idea. We do need rules, and people to applythem, for the good of patients and the studiesthemselves.Years ago, we had few regulations, and

Masterpiece

“It’s not so important that information is missing.

It’s important that what you have is good”

that was wrong. But when you go to thebureaucrats, you always seem tobe fight-ing a losing battle. You ask them, ‘Whatdo these words mean?’ They have oneinterpretation. I have another. A thirdperson has another. What can you do?”She tentatively suggests that one centralEuropean committeemight help, so thatseparate authority wouldn’t have to besought from regulators in each country:but EU regulations are not famous fortheir clarity.

IT’S ABOUT PATIENTSDespite all this, Valagussa is a greatbeliever in international, multicentretrials. They bring benefits to a far largergroupof patients than single-centre stud-ies. “I think you have to believe you aredoing your best for patients, and to share the optionsyou have for treatment in your country with othercountries.More patients benefit if you have severalsites, working as if they are one specialised centre.You get a good exchange of information betweeninvestigators, and the focus is on improvement.”Thepatient, she emphasises, shoulddrive every-

thing. It’s important to work with them before,during and after trials, often through patient organ-isations.User input into thedesign of consent formsis particularly important, she says. It is too easy todesignconsent forms that only cliniciansunderstand– and even they sometimes find them difficult.Iwonderwhether her background as a nurse has

helped provide a patient-conscious counterpoint tothe perspective of doctors in designing trials. Sheshrugs. Not really, she says. And as we begin totouchonher personal contribution andqualities, theanswers begin to dry up. I learn that she is single,sees a great deal of her 10 nephews andnieces, andtheir 10 children, and likes travelling, readingthrillers and listening to classical music. But shedoesn’t wish to go into details aboutwhatmakes her

tick. “If you let me talk about protocols, that’s fine.Otherwise, I stay quiet.”Actually, what motivates her has become obvi-

ous aswe talked about herwork, and about the debtshe feels to Bonadonna for his confidence in hersince her earliest days at the Institute. “I’ve alwaysappreciated that I’ve been able to talk openly withall the clinicians I’veworkedwith – just sensing thatwe were, and are, a team, working together out ofscientific curiosity.We all have this same challengeaheadof us, framing the clinician’s perspective in therightway so thatwe can test whatwe think accord-ing to the correct methodologies.”And sometimes,when theharddata showsome-

thing really exciting, the shy person who wants tokeep the personal out of the professional can’t helpacknowledgingher personal investment in thework.“When you start a study, yourmain priority has to benot to harm our patients for the sake of a scientificidea. Butwhen you get the initial results, sometimesyou cannot help being excited.You get a leap inside,and say: yes, we are on the right road.We have notsolved it, but we are on the right road.”

“You have to believe you are doing your best for patients,

and share the options you have with other countries”

Masterpiece

An internationalist. Valagussa goes out of her way to help new countries and newinstitutions participate in multicentre trials, even though this complicates her taskof controlling the quality of the data collected. She is pictured here at a regionalbreast cancer conference in Uruguay, 1999

ImpactFactor

Hypoxia modification withradiotherapy for bladder cancer

�Mary Gospodarowicz

Theoutcomes in bladder cancer treatedwith radiotherapy are suboptimal. Recently,Hoskin et al.

reported improved survival in patients with bladder cancer treated with radiation therapy with

concurrent hypoxia-modification therapy. These results are promising butmust be viewed in the

context of previous studies and alternative treatment approaches.

T hemanagement ofmuscle-inva-sive bladder cancer continues tobe controversial; there is no solid

evidence that overall survival hasimproved in the past two decades. Themost commonapproach for bladder can-cer management is radical cystectomy;bladder-conserving approaches withradiotherapy are far less frequently used.The major challenge in the latterapproach are the limitations of radio-therapy due to the proximity of dose-limiting organs including the bladder,rectum and adjacent small bowel.A number of approaches have beendeveloped using radiation-sensitising

chemotherapy to improve local control.1

In addition, hypoxia is present in mostsolid human tumours and attempts toovercome its effect have been tried inhead and neck cancer and, to a lesserextent, bladder cancer.2,3

In a recently publishedpaper,Hoskinet al.4 report the results of a prospectiverandomised trial comparing the efficacyof external-beam radiotherapywith andwithout concurrent carbogen andnicoti-namide. This trial was based onpreviousphase II trials that used a similarapproach, that is, hypoxiamodification.In the present trial,4 333 patients wererandomly assigned to either radiotherapy

alone (55Gy in20 fractions over 4weeksor 64Gy in 32 fractions over 6.5weeks)or radiotherapy with concurrent carbo-gen (2%carbon dioxide and 98%oxygenat 15 l/min for 5 min before and duringradiotherapy) and oral nicotinamide(60 mg/kg administered 1.5–2 hoursbefore each fraction of radiation).Patients were stratified by centre andtheir characteristics werewell balancedexcept for a slightly higher proportion ofpatients with T3 tumours (23.9% vs18%) in the radiotherapy-alone arm.Theprimary endpoint of the study wascystoscopic local control at six months.The secondary endpoints included the

This article was first published inNature Reviews Clinical Oncology 2011 vol.8 no.3, and is published withpermission. © 2011Nature Publishing Group. doi:10.1038/nrclinonc.2011.5, www.nature.com/nrclinonc

ImpactFactor

overall survival rate and local-relapse-freesurvival. The results showed improvedoverall survival at five years in the cohortreceiving radiotherapy, carbogen andnicotinamide (50% vs 39%).4

The results of this trial are surprisingbecause no previous phase III trial ofradiotherapy in bladder cancer hasshowed improved overall survival,including large randomised trials. Theresults are difficult to interpret becausethe primary trial endpoint of improvedcystoscopic local control was notachieved. In bladder cancer trials, theuse of local tumour control as an end-point is problematic because manypatients do not have follow-up cysto-scopies owing to the development ofprogressive distant disease or comor-bidities.5 Moreover, the higher rateof salvage cystectomy in the radiother-apy-alone group suggests improved localcontrol in the combination arm (13 vs 23salvage cystectomies).However, in termsof survival, the higher rate of salvagecystectomy should have compensatedfor the lower local control in radiother-apy-alone patients. The cohort treatedwith radiotherapy alonehadmoredeathsunrelated to bladder cancer (29 vs 24patients); therefore, it is difficult to becertain that the treatment interventioncaused the overall survival benefit. Thetrial usedCT todefine clinical target vol-umes but no details were providedregarding the verification of treatmentdelivery.4

There are considerable problemswith the reproducibility of radiotherapydelivery to bladder cancer that aremostlyrelated to variation causedbybladder fill-ing between fractions and during treat-ment.6 Trials of image-guided andadaptive approaches to overcome thisproblem are ongoing. The difficulties inimaging the actual tumour, rather than

the bladder, pose additional problemsthat some investigators have tried toovercome by injecting lipiodol aroundthe tumour to facilitate real-time imageguidance.7,8

Themajority of patientswithmuscle-invasive bladder cancer are managedwith radical cystectomy and pelviclymph-node dissection,while thosewhoare poor surgical candidates are referredfor external-beam radiotherapy.A num-ber of investigators have tried to popu-larise bladder conservation strategiesbasedoncombinedmodality approacheswith concurrent chemotherapy andradiotherapy. Unfortunately, littleprogress has been made in this area inthepast twodecades. Trials performed inthe 1980s and 1990s showed superior-ity of concurrent cisplatin and radio-therapy when compared withradiotherapy alone.9,10 However, the sin-gle-agent cisplatin had no impact ondistant-metastasis rate and, therefore, nosurvival advantage.10 Studies of adjuvantmultiagent cisplatin-based chemother-apy showed a very modest survivalimpact.9A surgical approach is generallypreferred as it defines the microscopicdisease extent in the primary tumourand regional lymph nodes. Radical cys-tectomy offers improved local controlfor tumours confined to the bladder,and pelvic lymph-node dissection hasbeen shown to cure a proportion ofpatients with involved pelvic lymphnodes.Unfortunately, theprice is the lossof natural bladder function, and althoughmodern continent diversion techniquesoffer improved quality of life, the long-termeffects of surgicalmanagement areimperfect.9

Bladder preservation strategies aremuch more complex than the standardsurgical approach and require closecooperation betweenurologists and radi-

ation oncologists with regards to patientselection, response assessment andongoing management. The best candi-dates for bladder preservationhave smallT2 tumours with no coexistent carci-noma in situ.5 Optimal survival has beenachieved with immediate salvage cys-tectomy in patients who do not achievelocal control or relapse owing tomuscle-invasive disease. The small proportion ofbladder cancer patients considered forradiotherapy hinders clinical trials inthis area. Most studies are small andrequire a prolonged accrual phase. Theapproach in theHoskin et al.4 trial mer-its attention and further study, espe-cially in patients not fit for moreaggressive approaches. However, as theHoskin et al.4 study shows, distant fail-ures continue to be a major problem inthis group of patients. It is important tonote that the approach taken does notaddress the issues of micrometastaticdisease and, therefore, is unlikely tohave a major impact on overall survival.

Details of the references cited in this article can

be accessed at www.cancerworld.org

Practice points� Bladder cancer management

remains a challenge for radiationoncologists; attention to patientselection, optimal treatment plan-ning and delivery is important

� There is a need for studies ofadaptive radiotherapy deliveryapproaches

� The role of hypoxia modificationmerits further study

� Participation in clinical trials ofcombinedmodality approaches isencouraged

Author affiliations: Princess Margaret Hospital, University of Toronto, Canada

ImpactFactor

An 18-gene signature (ColoPrint)for colon cancer prognosis

� Iain Tan and Patrick Tan

ColoPrint is an 18-gene expression signature designed to predict disease relapse in patients with

early-stage colorectal cancer (CRC).We discuss the potential impact of ColoPrint on clinical

practice, and its contribution to our knowledge of CRCmolecular heterogeneity.

Manyoncologists are familiarwithMammaPrint (Agendia, Ams-terdam, TheNetherlands) and

OncotypeDX(GenomicHealth,RedwoodCity, CA, USA), two multi-gene assaysused to predict disease relapse and guideadjuvant therapydecisions inpatientswithearly-stage breast cancer. Recently, bothcompanies have published gene-expres-sionclassifiers forpredictingdisease relapsein early-stage colorectal cancer (CRC).1,2

Here,wediscuss thepotential clinical andscientific impact of one of these classifiers–ColoPrint (Agendia).CRC is the third leading cause of

global cancer mortality. Outcomes forpatients with early-stage CRC are het-erogeneous, with five-year survival ratesranging from 72% to 83% in stage IIdisease and from44% to83% in stage IIIdisease.3 In the past two decades, ran-domised trials have demonstrated a sur-vival advantage for patients treated withsurgery andadjuvant chemotherapy,4 par-

ticularly those with stage III disease.However, in these trials, many patientswere cured by surgery alone, suggestingthat it might be possible to omitchemotherapy in selectedpatients.Clin-ical guidelines currently recommendobservation for stage I disease and adju-vantchemotherapywithacombinationofa fluoropyrimidine and oxaliplatin forthosewith stage III disease.In stage II CRC, the benefit of adju-

vant chemotherapy is contentious, withASCOrecommending the integrationofclinical risk criteria to select patients foradjuvant therapy.5 Identifying molecularmarkers that can informtherapeuticdeci-sions, suchas theneed for treatment andtype of adjuvant therapy, would betremendouslyuseful.Toaddress thischal-lenge,Salazar et al.1 analyzed fresh-frozentumour tissues from 188 patients withstage I to IV CRC using Agilent gene-expression microarrays. By correlatingtheexpressionofmore than40,000genes

with metastasis-free survival, they iden-tified anoptimal set of 18 genes thatwasused to construct the ColoPrint prog-nostic classifier. In an independent vali-dation series of 206 patients with stage Ito III CRC, 60% of patients were classi-fied as ‘low risk’, with a five-year relapse-free survival (RFS) rate of 87.6%. Theremaining40% ‘high-risk’patients exhib-ited a RFS rate of 67.2% (HR=2.5; 95%CI1.33–4.73;P=0.005). Inmultivariateanalyses,ColoPrint remainedoneofmostsignificant prognostic factors (HR=2.69;P=0.003), and instage IICRC,ColoPrintwas superior to theASCOcriteria for theassessment of cancer recurrence risk(HR=3.34; P=0.017). The authors con-cluded that, comparedwithconventionalclinicopathological criteria alone, Colo-Print provides more accurate informa-tion on the risk of recurrence and mayfacilitate selection of low-risk patientswho can be spared chemotherapy.While these results are encouraging, it

This article was first published inNature Reviews Clinical Oncology 2011 vol.8 no.3, and is published withpermission. © 2011Nature Publishing Group. doi:10.1038/nrclinonc.2010.229, www.nature.com/nrclinonc

ImpactFactor

is prudent to interpret them in thecontextof an early discovery study. Several gene-expression classifiers for predictingCRCrelapse have been described,6,7 but nonehave achieved clinical utility. It is worthnoting that studies relying on fresh-frozentissue (for example ColoPrint) typicallyhavemodest sample sizes andcannotben-efit from archival material collected fromrandomisedclinical trials.Asacomparison,Oncotype DX (colon), the parallel CRCprognostic classifier developed using for-malin-fixed paraffin-embedded tissues,was tested in more than 1800 patientsfrom four adjuvant trials.2 Therefore, fur-ther retrospective validation of ColoPrintin large independent cohorts is clearlyrequired.Fortunately, a prospective study,PARSC(Prospective study for theAssess-ment of Recurrence risk in Stage II Col-orectal patients using ColoPrint) hasalreadybeen initiated to evaluate theper-formanceofColoPrint in theclassificationof patients in the clinical setting.8

The potential for gene signatures toinfluence treatment decisions dependson the disease stage. Molecular markersaremost likely to impact themanagementof stage II disease, where the need foradjuvantchemotherapy is alreadybasedonassessment of clinical risk features.Mol-ecularly, microsatellite instability (MSI)status is a marker of good prognosis inpatients with stage II CRC and may beassociated with a lack of benefit fromadjuvant fluoropyrimidine therapy.9

Indeed,mostMSIhigh (MSI-H)patientswere identified as ‘low risk’ by ColoPrint.However, the 48% discordance observedbetweenColoPrint and theASCOclinicalrisk criteria1 suggests an additional dis-criminative value of ColoPrint beyondclinical characteristics. Stage II patientsidentified as ‘low-risk’byColoPrint exhib-ited an excellent five-year survival similarto that seen for stage I disease, raising thepossibility that ColoPrint may identify

stage II patients forwhomchemotherapycan be avoided. That said, we mustremember that good prognosis does notnecessarily mean lack of benefit fromadjuvant therapy. For example,OncotypeDX (colon)hasnot been shown tobepre-dictive in stage II CRC, despite its prog-nostic significance.2 Further studies shouldalso be performed to establish if Colo-Print is purely prognostic or whether it ispredictive of treatment benefit as well.In stage IIICRC,adjuvantchemother-

apy is the standardof care. Inour opinion,oncologists are highly unlikely to omitchemotherapy altogether in medically fitpatientswithstage IIICRCunless thedatasupporting excellent prognosis in molec-ularly low-riskpatients is very compelling.The studybySalazar et al.1 cannot addressthe role of ColoPrint in stage III disease,since there were only 62 patients withstage III disease and there was a trendtowards inferiorRFS inhigh-risk patients(P=0.1). Nevertheless, a validated prog-nostic signature for stage IIICRCpatientsmight still be useful to identify low-riskpatients forwhomoxaliplatin chemother-apy might be omitted and who might betreated with a fluoropyrimidine alone.Moreover, with the exception of oxali-platin, stage III CRC has demonstratednotable failures for drugs thatwere effica-cious in the metastatic setting, such asbevacizumab, cetuximab and irinotecan.Given the curative intent of treatment instage IIICRCandthevast investment intothese completed trials, itmight be fruitfulto search formolecularmarkerspredictiveof selective benefit for therapies that oth-erwise do not provide an advantage in anunselected population.The present study also broadens our

knowledge regarding the inherentmolec-ular heterogeneity ofCRC.1Usingunsu-pervised clustering techniques, threemolecular subgroupswere identified thathad different survival outcomes. These

groups were differentially enriched forBRAF activatingmutations andMSI-H,suggesting unique underlying biologies.Notably, only the largest subgroup(n=110) was used to develop the prog-nostic signature. Given the distinct bio-logicalmakeupof these three groups, it isplausible that the prognostic impact ofColoPrint is specific to the biologicalsubgroup from which it was developed,analogous to thequestionableprognosticvalue ofOncotypeDX inHER2-positivebreast cancer.10 Salazar et al.1 do not pro-videprognostic informationofColoPrintin the three biological subgroups – thisshouldalsobeaddressed ina future study.Investigations addressing the relation-ship of ColoPrint to other molecularmarkers (for example, 18q loss of hetero-zygosity,KRASmutation status andCpGisland methylation subtypes) are alsowarranted.Inconclusion,Salazar andcolleagues

are tobecommended for their promisingfindings that ColoPrint might provideadditionalprognostic informationbeyondclinicopathological criteria in early-stageCRC.Weeagerly await the results of theongoing clinical trial seeking to prospec-tively validateColoPrint.

Details of the references cited in this article can be

accessed at www.cancerworld.org

Practice pointIn colorectal cancer, novel molecularmarkers such as gene-expression signa-tures offer the potential of improvinguponcurrentprognosticmodels that arebasedonclinical criteria.However,wide-spread acceptanceof thesemarkerswillnecessitate identifying opportunitieswhere they directly influence clinicalmanagement decisions.

Author affiliations: Department of Medical Oncology, National Cancer Centre, Singapore (I Tan). Cancer and Stem Cell Biology, Duke–NUS Graduate Medical School, Singapore (P Tan)

NewsRound

N E W S R O U N DSelec ted repo r t s ed i t ed by Jane t F r i cke r

Treatment by gynaecologiconcologists improves outcomein endometrial cancer� Journal of Clinical Oncology

Patients with endometrial cancer treated bygynaecologic oncologists weremore likely

to undergo staging surgery and receive adju-vant chemotherapy in advanced cancer, a USstudy has found. The researchers also reporteda higher five-year disease-specific survivalamong endometrial cancer patients with stageII–IV diseasewhowere treated by gynaecologiconcologists.

Over the last decade the number of annualdeaths fromendometrial cancer has doubled intheUS. Studies of patientswith ovarian cancerhave shown that those receiving care fromgynaecologic oncologists underwent morethorough staging surgery and received morechemotherapy in high-risk disease. This, in turn,led to improved survival outcomes. Despite alack of evidence for survival benefit for patientswith endometrial cancer, the American Col-lege ofObstetrics andGynecology recommendsthat these patients should also be referred togynaecologic oncologists.

In the current study, John Chan and col-

receive chemotherapy for advanced disease,with 22.6% in groupA receiving chemotherapyversus 12.4% in group B (P<0.001).

For women with stage II–IV disease, thefive-year disease-specific survival (DSS) of groupA patients was 79% versus 73% for group B(P=0.001). For stage III–IV disease, women ingroupAhada five-yearDSSof 72%versus 64%in group B (P<0.001). However, no associationwith DSS was identified among women withstage I cancers. On multivariable analysis,younger age, early stage, lower grade, andtreatment by gynaecologic oncologists wereall found to be independent prognostic factorsfor improved survival.

“Directed care by gynecologic oncologistswas associated with more extensive lymphnode resection and subsequent adjuvant ther-apy. Most importantly, care provided by gyne-cologic oncologists improved the survival ofthosewith high-risk (stages II to IV, grade 2 and3, and high-risk histologies) disease,” write theauthors, adding, that to their knowledge this isthe first population-based study to haveanalysed the impact that gynaecologic oncol-ogist care has on endometrial cancer patients.

Nearly 80%of endometrial cancer patientsin the study, stress the authors, did not receivecare from gynaecologic oncologists. “Clearly,further research is needed to identify the

leagues from theUniversity of California at SanFrancisco, undertook todetermine the influencethat care by gynaecologic oncologists had onboth the type of treatment and the survival ofpatients with endometrial cancer. Between1988 and 2005 the investigators obtained datafrom theMedicare data bases and from theSur-veillance, Epidemiology and EndResults (SEER)programme. The speciality of the treating sur-geonwas foundby linking their uniqueprovideridentification numbers in the data bases toinformation collected by theAmericanMedicalAssociation. Kaplan–Meier and Cox propor-tional hazardmethods were used for analyses.

Results show that of the 18,338 womenidentified with endometrial cancer in the databases, 21.4% received care from gynaeco-logic oncologists (defined as group A); while78.6% were treated by other clinicians(defined as group B).

Women in group Awere older, with 49.6%in group A aged over 71 years compared to44% ingroupB (P=0.001); theyhadmore lymphnodes removed, with 22% in group A havingmore than 16 nodes removed compared to17% in group B (P<0.001) they presentedwithmore advanced cancer, with 21.9% in group Ahaving stage III–IV cancers versus 14.6% ingroup B (P<0.001); they had higher-gradetumours (P<0.001) and theyweremore likely to

NewsRound

disparities in endometrial cancer treatment andpotential barriers to accessing subspecialtycare,” they write.

Limitations to the study included a lack ofinformation on the extent of residual diseaseafter cytoreductive surgery for advanced dis-ease, lack of central pathology review, unknowntypes and cycles of adjuvant chemotherapyandunspecified treatment for recurrent disease.“Without central pathology review, it is possiblethat a change in gradeof diseasemayaffect theresults of this study,” write the authors.

� J K Chan, AE Sherman, DS Kapp et al.

Influence of gynecologic oncologists on the

survival of patients with endometrial cancer. JCO

doi 10.1200/JCO.2010.31.2124, published online

24 January 2011

Choice of surgeon isa significant influenceon outcome in DCIS� JNCI

The most important determinants of out-comes forwomenwith ductal carcinoma in

situ (DCIS) are associatedwith tumourmargins,whether or not they received radiotherapy,whether or not they underwent mastectomy,and the treating surgeon, a retrospective USstudy has found.

“An important implication of our work isthat surgeonsmayplay a critical role both in thesurgical treatment choices made by patients(and in the receipt of radiation therapy). Becausethese are the most important factors in pre-dicting outcomes the substantial variation bysurgeon suggests that the quality of DCIS carecould be improved,” write the authors AndrewDick and colleagues fromtheRANDCorporation(Pittsburgh, PA), a non-profit-making institu-tion,with the remit to improve policy and deci-sionmaking through research and analysis.

The goal for treatingDCIS, or non-invasivebreast cancer, is to reduce the likelihood ofdeveloping invasive breast cancerwhile respect-

ing patient preferences for treatment options,which include breast conserving surgery alone,breast conserving surgery followedby radiation,and mastectomy. Since DCIS is non-lethal,physicians’ attitudes regarding optimal man-agement andpatient preferencemayplay a rolein treatment decisions.

To determine the comparative effectivenessof treatment strategies, and identify key factorsassociatedwith variations in outcomes, Dick andcolleagues conducted a retrospective study of994womendiagnosedwithDCIS between1985and 2000. The investigators identified subjectsthrough two large tumour registries: theMon-roe County (New York) tumour registry, andthe tumour registry at the Henry Ford HealthSystem in Detroit. Margins were defined aspositive (when cancer cells extended to theedge of the resected tissue); negative (whencancer cells were more than 2 mm away fromthe edge of the tissue); or close (when cancercells were present within 2mm of the edge).

Results showed that the overall differencesin predicted five-year disease-free survival rateswere 0.993 for mastectomy, 0.945 for breast-conserving surgery with radiation therapy and0.824 for breast-conserving surgery withoutradiation therapy,with all the differences foundto be statistically significant (Pdiff <0.001 foreach of the differences). Similarly, each of thedifferences at 10 years was statistically signifi-cant (P<0.001).

In all the treatment groups, except breastconserving surgery without radiation therapy,the rates of recurrencewere statistically signif-icantly different according to margin status,with positive margins found to be associatedwith substantially higher recurrence rates.Furthermore, variation by surgeon accountedfor 15%–35% of the subsequent ipsilateralfive-year recurrence rates, and for 13%–30%of10-year recurrence rates.

“Although variation by surgeon could begeneratedbypatients’ preferences, the extent ofvariation and its contribution to long-termhealth outcomes are troubling,” write theauthors, adding that furtherwork is required todetermine why women with positive marginsreceive no additional treatment andwhymar-

gin status and receipt of radiation therapy varyby surgeon. Additionally, they add, there is cur-rently no consensus onwhat constitutes a neg-ative margin.

In an accompanying commentary, BethVirnig and Todd Tuttle of theUniversity ofMin-nesota askhowpatients shouldgoabout select-ing health providers, in the knowledge that upto 35%of the variation in outcomes is based ontheir choice of physician, but that there are noactionable characteristics that can be takeninto account.

One solution, they suggest, would be topublish the scores for all physicians perform-ing breast cancer surgery in a particular area.“With this approach, it would not matter whyone physician had higher or lower recurrencerates or positive margin rates, the rates wouldsimply be reported so that women could takethis into account when selecting a provider,”they write.

The challenge, they add,will be for the pro-fessional community to identify factors that areassociatedwith theunexplained physician vari-ability and to use that information to promoteidentification of high-quality providers or qual-ity improvement activities.

� AW Dick, MS Sorbero, GM Ahrendt et al.

Comparative effectiveness of ductal carcinoma in

situ management and the roles of margins and

surgeons. JNCI 19 January 2011, 103:92-104

� B Virnig, TM Tuttle. Random physician effect

and comparative effectiveness of treatment for

ductal carcinoma in situ. ibid, pp 81–82

Protocols needed to cutdelays in giving antibioticsfor febrile neutropenia� British Journal of Cancer

Astudy providing detailed insights into themanagement of chemotherapy-induced

febrile neutropenia in a UK Cancer Networkshows that while the condition is generallyrecognised early and managed appropriately,

NewsRound

improvements are needed in the timely admin-istrationof antibiotics. Theprospective study, theauthors believe, highlights the need for intro-ducing network-wide clinical care pathwaysto improve outcomes.

Febrile neutropenia (the development offever in patients with abnormally low levels ofneutrophil granulocytes) is a complication ofchemotherapy associated with considerablemorbidity and mortality. The UK NationalChemotherapy Advisory Group (NCAG) – setup to provide advice on the delivery of high-quality chemotherapy services to the NationalCancer Director and Department of Health –have produced guidelines on febrile neu-tropenia, recommending ‘treat and transfer’arrangements (if hospitals do not have appro-priate facilities), and an arrival to deliverytime of antibiotic administration for neu-tropenic sepsis of less than one hour.

“The NCAG recommendations provide aframework for a process of assessment, decisionto treat, informed consent and prescription ofchemotherapy, and emphasise the importanceof detailed standardised consent forms and theinvolvement of senior trainedoncologymedicalstaff ,” write the authors of the study, SimonChowdhury and colleagues from Guy’s &St Thomas’ NHS Foundation Trust (London, UK).

In this study, Chowdhury and colleaguesundertook a prospective studyof all the cases ofchemotherapy-induced febrile neutropenia inthe South West London Cancer NetworkbetweenMay and August 2007. Data recordedat seven hospitals serving a population of1.4millionpeople includeddemographics, treat-ment histories, management of febrile neu-tropenia and outcomes.

Results showed that, in all, 71 admissionsfor febrile neutropenia were reported, involv-ing 64 patients, with seven patients admittedon two separate occasions. Fifty-nine per centof patients (n=38) were female and 41%(n=26)male, with amedian age of 60 years. Ofnote, one-third of patients with febrile neu-tropenia in the studywere older than 65 years,which the authors suggest supports the notion“that increasing age is an independent pre-dictor of development of febrile neutrope-

IMRT spares head andneck cancer patientsdry mouth symptoms� Lancet Oncology

Sparing the parotid glands of patients withhead and neck cancer by using intensity

modulated radiotherapy (IMRT) reduces the inci-denceof xerostomia (a drymouthdue to lackofsaliva), reports thephase III UKPARSPORT study.

While radiotherapy is themain non-surgi-cal treatment for squamous-cell carcinoma ofthe head and neck, radiation-induced xerosto-mia is a commonly reported late side-effect.Lack of saliva affects speech and swallowing,and can accelerate dental caries. In comparisonwith conventional radiotherapy, IMRT, whichallows focused radiation delivery to tumours,can reduce irradiation of the parotid glands.

In the current study,which took place at sixUK centres between January 2003 andDecem-ber 2007, Christopher Nutting and colleaguesfrom the Institute of Cancer Research (Sutton,UK), randomised 94 patients with histologi-cally confirmed squamous carcinoma (T1–T4,N0–N3,M) to parotid sparing IMRT (n=47) or toconventional radiotherapy (n=47). In bothgroups, the primary tumour and involved lymphnodes were treated with 65 Gy, delivered in 30daily fractions, five days aweek. The investiga-tors undertook measurement of salivary flowprior to radiotherapy, at week 4 of treatment,and then at 2 weeks and 3, 6, 12, 18 and 24months after radiotherapy.

Results at 12months show that grade 2 orworse xerostomia symptomsoccurred in74%ofpatients given conventional radiotherapy versus38% of patients given IMRT (P=0.0027). At 24months, grade 2 or worse xerostomia occurredin 83% of patients given conventional radio-therapy versus 29%given IMRT (P<0.0001).

At 12 months, unstimulated saliva flowfrom the contralateral parotid glandswasnotedin 47% of patients treated with IMRT versusnone treated with conventional radiotherapy(P<0.0001), while at 24 months unstimulatedsaliva flow occurred in 44% of patients in the

nia.” The seriousness of the condition wasunderlined by the fact that three of thepatients (4.2%) died as a direct consequence ofneutropenic sepsis.

Forty-five patients (63%) were admitteddirectly to a specialist oncologyorhaematologyward,while 21 (30%)were seen first in the acci-dent and emergency departments. The mediantime from arrival to nursing assessment was10minutes (range0–135mins), and themediantime to first assessment by a clinician was40 minutes (range 0–230 mins). The mediantime fromarrival to administrationof anantibi-oticwas135minutes (range15–550mins),withonly 9 out of 50 patients receiving antibioticswithin 60minutes.

“Our study has provided an important anddetailed insight into the incidence andmanagement of chemotherapy-induced febrileneutropenia in a representative cancer net-work in theUnitedKingdom,”write the authors,adding that the area that most needs to beaddressed is the time interval between arrival atthe hospital and treatment.

“To achieve this, physician and nursingprotocols to standardise and streamline clinicalcare pathways for thewhole network are underconsideration,”write the authors, adding that itis hoped that the recommendation for NICE toprovide anationwide policy formanagement ofneutropenic sepsiswill lead to the introductionof a standardised approach both within andacross networks.

One issue, add the authors, is that periph-eral hospitalsmay not be staffedwith 24-houroncology services, making it crucial that thesesites have access to well-designed protocolsand expert consultant advice. Furthermore,patient education regarding what to do in theevent of a chemotherapy-induced complicationis fundamental to ensuring people receiveprompt appropriate care.

� M Okera, S Chan, U Dernede. A prospective

study of chemotherapy-induced febrile neutro-

penia in the South West London Cancer Network.

Interpretation of study results in light of

NCAG/NCEPOD findings. Br J Cancer 1 February

2011, 104:407–412

NewsRound

IMRT group versus none in the conventionaltherapy group (P=0.0068).

The only recorded acute adverse event ofgrade 2 or worse that differed significantlybetweenthe treatmentgroupswas fatigue,whichoccurred in 41% of patients given conventionalradiotherapy versus 74%given IMRT (P=0.0015).Significant differenceswere also noted in stimu-lated saliva flow fromthecontralateral parotid at12 months (P<0.0001). The estimated two-yearoverall survivalwas76%withconventional radio-therapy versus 78%with IMRT.

“Our trial has shownaclinically and statisti-cally significant reduction in xerostomia,improved salivary flow, and improved quality oflife, and this strongly supports a role for IMRT inhead and neck squamous cell carcinoma,” con-clude the authors, adding that the results of thePARSPORT trial are likely tobegeneralisable toallhead andneck tumours forwhich conventionalradiotherapy is used. It is possible, they say, thatfurther reductions in severe xerostomia couldbeachieved by additionally sparing the sub-mandibular andmucosalminor salivary glands.

Fatigue was unexpectedly found to begreater in patients treated with IMRT. This, saythe authors, could be due to greater radiationdoses being delivered to non-tumour tissues.

One limitationof the study, they add, is thatit was not possible tomask the treatment frompatients or clinicians due to the differences intreatment delivery. “However, results that relateto multiple secondary endpoints support theprimary analysis and the size of the observedeffect is unlikely to bedueentirely to assessmentor reporting bias.”

In an accompanying commentary, AndyTrotti from the Moffitt Cancer Center (Tampa,Florida) andAvi Eisbruch from theUniversity ofMichigan Medical Centre (Ann Arbor) say thatfuture work should systematically explore theprioritisation of different components of thesalivary gland system, since a clinical benefit ofsparing the submandibular glands can beobtained over the parotid glands. “The parotidglands provide watery saliva during eating,which is largely replaceable by consumingmorewater or lubricants. The submandibular, sublin-gual, and minor salivary glands provide muci-

nous saliva, associatedwith the resting sense ofmoisture and dry mouth symptoms.” Furtherpossibilities, they add, include gland repair orregenerative strategies with stem cells,acupuncture, or acupuncture-like stimulation.

� CM Nutting, JP Morden, KJ Harrington et al.

Parotid-sparing intensity modulated versus

conventional radiotherapy in head and neck cancer

(PARSPORT): a phase 3 multicentre randomised

controlled trial. Lancet Oncol February 2011,

12:127–136

� A Trotti, A Eisbruch. Reducing xerostomia

through advanced technology. ibid pp 110–111

Cannabis improvescancer patients’ appetitesand sense of taste� Annals of Oncology

Theactive ingredientof cannabis improved theappetite and sense of taste of patients with

advancedcancer, buthadnoeffecton their calo-rie intake, aUSproofofprinciple studyhas found.

Loss of appetite is common among cancerpatients, bothdue the cancer itself and to treat-ment affectingpeople’s senseof taste and smell.As a result, theyexperiencedecreasedenjoymentof food, which in turn can lead to weight loss,anorexia, reduced quality of life and decreasedsurvival. For a long time health professionalsthought nothing could be done and advisedcancer patients to ‘cope’ with chemosensoryproblems by eating bland, cold and odourlessfood. More recently, however, the potential ofdelta-9-tetrahydrocannabinol (THC) – themainpsychoactive ingredient in cannabis, which isthought to increase appetite via endocannabi-noid receptors – has been recognised. Studieshave shown that THC increases appetite in ani-mals, healthy humans and patients withacquired immunodeficiency syndrome (AIDS),but its ability tohaveaneffect in cancer patientshas not been consistently reported.

In the current study – conducted in twoCanadiancancer centres inEdmontonandMon-

treal –WendyWismer and colleagues from theUniversity of Alberta (Edmonton, Canada)hypothesised that THCmay favourably alter thechemosensory perception of cancer patients.

In the randomisedplacebo-controlledphaseII double-blind pilot study, which took placebetween May 2006 and December 2008, 21patientswith advanced cancer (excluding braincancers)whohad been eating less as a result oftheir illness for two weeks or more, were ran-domised to receive THC (n=11) or to placebo(n=10). The active capsules contained 2.5mgofTHC, with patients asked to take them once aday for the first three days, then twice a daythereafter, with the option to increase the doseto amaximumof20mgaday. Treatment ran for18days. Questionnaireswere conducted before,during and at the end of the trial.

Fromquestionnaires, researchers found that73% of THC-treated patients reported anincreased overall appreciation of food com-pared with 30% of patients receiving placebo,and that 55% said that the medication “madefood taste better”, compared with 10% takingplacebo (P=0.04). Half of the patients whoreported odours to beunpleasant at baseline nolonger found odours offensive with THC treat-ment (P= 0.083).

Although no difference was found in thetotal number of calories consumed by the twogroups, the THC-treated patients tended toincrease the proportion of protein they ate,and 55% reported that savoury foods tastedbetter,whereas nopatients in theplacebogroupreported an increased liking for these foods. Inaddition, THC-treated patients reported betterquality of sleep (P=0.025) and relaxation(P=0.045) in comparison the placebo group.

“Our pilot study demonstrates that THC,comparedwithplacebo, improvedandenhancedchemosensoryperception, alteredmicronutrientpreference, appeal of savory foods, appetite,relaxation, and quality of sleep for advancedcancer patientswith chemosensory alterations,”write theauthors, adding that thedatawill assistin the development of larger phase II trials byfacilitating sample size calculations.

Inevitably, they add, questions are raisedabout theability toblindTHCtreatmentbasedon

NewsRound

its well-known psychoactive characteristics.“However, the timedadministrationof lowdosesand the lackofdifferences inAEs [adverseevents]between treatment groups suggest that thisproblemwas likelyminimal,” write the authors.

� TD Brisbois, IH deKock, SM Watanabe et al.

Delta-9-tetrahydrocannabinol may palliate altered

chemosensory perception in cancer patients: results

of a randomized, double-blind, placebo-controlled

pilot trial. Ann Oncol doi:10.1093/annonc/mdq727,

published online 22 February 2011

Study shows occultmetastases have littleeffect on outcome� New England Journal of Medicine

Noadditional clinical benefit is obtained byevaluatingwomenwith breast cancer for

occult metastases, a US study has concluded.The National Surgical Adjuvant Breast andBowel Project (NSABP) B-32 trial found that thedetection of occult metastases could not beregarded as a discriminatory predictor of can-cer recurrence.

Clinicians have debated for some timewhether breast cancer patients whose lymphnodes initially test negative for disease, butwho have occult metastases (detected afterfurther evaluation), are at greater risk of recur-rence andmight benefit frommore aggressivetreatment. The standard approach, endorsedby ASCO and the American College of Sur-geons, is to slice sentinel nodes at 2.0-mm

intervals and stain samples with haematoxylinand eosin (H&E). However, controversy contin-ues over whether pathologists should be look-ing at more frequent intervals in order to findhidden cancers.

DonaldWeaverandcolleagues fromtheUni-versity of Vermont College ofMedicine (Burling-ton,USA) undertook theB-32 trial inwhich5611womenwithnoclinical evidenceofmetastaticdis-ease in thearmpitwere randomlyassignedtosen-tinel lymphnodebiopsyplus axillarydissectionorsentinel lymph node biopsy alone.

In the current sub-study, the 3887 patientsinwhommetastaseswere not detected under-went further evaluation. Tissue blocks from thenegative sentinel nodes were sent to a centrallaboratory for evaluation of additional sectionsthat were 0.5–1.00 mm deeper in the blockrelative to the original surface. The deeperanalysis included routine use of (H&E) testingand immunohistochemical staining.

Occultmetastaseswere detected in 15.9%of the patients whose initial sentinel nodebiopsy tested negative for cancer. Isolatedtumour cells (≤0.2mm) accounted for 11.1%ofthe metastases followed by micrometastases(>0.2–≤2.0 mm, 4.4%), and macrometastases(>2.0 mm, 0.4%).

Log-rank tests indicated a significantdecrease in overall survival (P=0.03); disease-free survival (P=0.02); and distant-disease-freeinterval (P=0.04) between patients in whomoccultmetastaseswere detected andpatients inwhom occult metastases were not detected.

However, the five-year Kaplan–Meiersurvival estimates for overall survival were94.6% for patients inwhomoccultmetastaseswere detected versus 95.8% for patients in

whomoccultmetastaseswere not detected; fordisease-free survival they were 86.4% versus89.2%; and for distant-disease-free survivalwere 89.7% versus 92.5%.

Themultivariable analysis identified severalother factors that influenced outcomes: olderage and larger primary tumour size adverselyaffected outcomes, whereas systemic chemo-therapy, endocrine therapy and radiationtherapy significantly improved outcomes.

Perhaps the most interesting interaction,say the authors, was with endocrine therapy,indicating that occultmetastases are associatedwithoestrogen-receptor-positive tumours (con-sidered a favourable prognostic factor) andthat endocrine therapy markedly reduces therisk of a poor outcome.

“Occult metastases were an independentprognostic variable in patients with sentinelnodes thatwerenegative on initial examination;however, the magnitude of the difference inoutcomeat five yearswas small (1.2 percentagepoints),” write the authors, adding that theirfindings argue against analysis of additional tis-sue levels or routine immunohistochemicalanalysis for sentinel-lymph-node evaluation.

One limitation of the study, they say, is thatno analysis would be able to detect all theoccult metastases.

Although thedifference in survival betweenwomen with and without occult metastasiswas small at five years’ follow-up, the investi-gators believe the study “warrants continuedfollow-up and analysis”.

� DL Weaver, T Ashikaga, DN Krag et al. Effect of

occult metastases on survival in node-negative breast

cancer. NEJM 19 January 2011, 364:412–421

Corrections and clarificationsISOPP’s foundersIn the cover story on Klaus Meier, published in the Jan-Feb issue, we reported thatMeier founded the International Society of Oncology Pharmacy Practitioners(ISOPP).We would like to clarify that it was Helen McKinnon of New Zealand whocame up with the idea of founding ISOPP in 1988, and she was elected ISOPP’s firstpresident in 1997. Meier was part of the board that initiated and developed the notionof ISOPP becoming an incorporated society in 1993, and he turned that notion intoreality, setting up ISOPP as an incorporated society under German law in 1996.

Myriad’s gene patentsIt was a federal district court that overturned some of Myriad’s BRCA patents inMarch last year, and not the Supreme Court, as was incorrectly reported in thearticle on Promoting genetic literacy, in the Jan-Feb issue of Cancer World. Myriad isnow appealing the decision in a hearing that started on April 4th. It’s chances ofsucceeding will have been dented by a re-evaluation of past policy conducted by theUS Justice Department, which has filed a brief asking the appeal judges to upholdparts of the ruling that overturned several of Myriad’s patents on the BRCA genes.

Cutting unnecessary deathsfrom cervical cancerCollective effort aims to narrow eleven-fold gap between worst and best in Europe

� Peter McIntyre

Given howpreventable cervical cancer is, setting up robust screening programmesmust feature

as a key element in Europe’s strategy for cutting deaths from cancer. But as this six country

initiative is finding out, it takes time, resources and attention to detail. Sharing experiences and

learning from themodel Finnish screening programme has been key to making progress.

It is ten years since the Euro-pean Union started to focusattention on fighting inequalitiesin cancer between countries,using the twin tools of compar-

ison of data and solidarity betweencountry programmes.Perhaps nowhere has that inequal-

ity been shown more clearly than incervical cancer, where incidence andmortality in some European countriesare five times higher than those withthe best organised screening pro-grammes. This translates into tens ofthousands of extra deaths of womenacross the whole of Europe, oftenwomen in middle age who are activeeconomically and key family members.What makes this tragedy the more

unacceptable is that cervical cancer isin most cases preventable or curable.

Despite the high profile that vaccinesagainst HPV infection have achieved,the missing ingredients are the old-fashioned public health virtues that goto make up population-based screen-ing. The gap is in planning, organisa-tion, training and perhaps politicalcommitment.There is also lack of knowledge and

a sense of distrust on the part of somewomen that inhibits them from goingfor check-ups.The net result is that women are

four times more likely to develop cervi-cal cancer over the course of a lifetime inEstonia, Lithuania or Slovakia than inFinland, while in Lithuania and Roma-nia they are eight to eleven times morelikely to die fromcervical cancer (Globo-can 2008 data – see box, p 60).The latest stage in a European pro-

gramme aimed at fighting cancerinequalities, EUROCHIP 3, waslaunched by the European Commis-sion in September 2008, with cervicalcancer as a major focus.The five countries officially

included in efforts to transformscreening systems are Bulgaria, Esto-nia, Latvia, Lithuania and Romania.They were chosen not simply becausethe figures were amongst the worst,but because teams of professionalswere already beginning to address theproblem, and there was something tobuild on. Poland, although notincluded in EUROCHIP 3, is workingalongside these countries to improveits own figures.Together, these countries consti-

tute a base for testing current knowl-edge on how to implement andmanage

Systems&Services

“Some women have screening much too frequently,

whereas some are underserved or never screened”

screening programmes to achieveacceptable coverage and quality stan-dards with medium or low levels ofhealthcare resources.

DEVELOPING EFFECTIVESCREENINGThe key planks for screening pro-grammes are that they invite women ina target age range (usually 30–59 yearsold) at regular intervals for high-qualityPap smears that are accurately read,and that they follow up womenwho donot attend or who have unusualsmears, and ensure high-quality treat-ment. To ensure the Pap smears are ofgood quality and accurately read,screening programmes also requiretraining for gynaecologists, generalpractice doctors (GPs), nurses and lab-oratory staff, as well as systematicmon-itoring and evaluation.

To close the gap in publicawareness and promoteatten-dance, public infor-mation and advocacy areneeded. Some of the countries on thislist do not even have a word for screen-ing in their own languages.Leading this EUROCHIP work

package is Ahti Anttila, who is theresearch director of the Finnish MassScreening Registry, which has be-come the system by which the rest ofthe world judges itself. Finland wasthe first country to institute a screen-ing programme, in 1962. No-one paidmuch attention until, in 1976, a paperin the American Journal of Epidemi-ology demonstrated that the incidenceof cervical cancer was 80% loweramongst women who had beenscreened than in the rest of thefemale population.

Although, themain inequalities in cer-vical cancer are between ‘old Europe’and the countries of central and easternEurope, Anttila points out that thereare also still inequalities within othercountries that lack a screening pro-gramme. “There are big countries likeGermany and Belgium where some ofthe target population is still missing.Some women have screening muchtoo frequently, whereas there could bea proportion, let us say about 20% ofthe target population, who are under-served or never screened.” It’s clearthat much more needs to be done toensure European countries adhere tothe current EU recommendations andguidelines on screening, he adds.

Systems&Services

A pilot screening programme in Cluj countyRomania. Left: Women of all ages queue upbeside the mobile screening unit in one of thecounty’s 356 villages. Below: Local GPs weretrained in carrying out smears and breastexaminations as part of the pilot; some of themhave now taken responsibility for this work, whileothers are still assisted by the mobile unit.Almost 80% of women in this region aged 25–64 years had never previously had a Pap test

EstoniaWith a population of 1.34 million,Estonia is the smallest country in theEUROCHIP group, and it was one ofthe first to get a screening programmeup and running, with nationwidescreening for women aged 30–59 start-ing in 2006. Estonia was motivated bya disturbing trend that saw the inci-dence of cervical cancer double in the30- to 49-years age group between theearly 1980s and 2000–2006.Pap smears are taken by trainedmid-

wives at 19 clinics around the country.However, despite efforts to organise thesystem, take-up has been disappointing,with only 15% of women attending.Meanwhile about 50% of the target

group of women have had private smeartests outside the screening programme.Anttila points out that, due to theabsence of a screening registry, it has notbeen possible to check on the quality ofthese smears or what follow-up treat-ment women have been offered.Epidemiologist Piret Veerus, from

theNational Institute forHealthDevel-opment in Tallinn, has overseen a studyto find out why women do not attend.She found that women wanted a

personal written letter inviting them toscreening and to be able to phone foran appointment at a time that suitedthem. Information levels were highamongst the Estonian majority, butfewer than half of the Russian minor-

ity even knew that a screening pro-gramme existed.Veerus would like to see a health

education programme in schools toalert young women to the risks of earlysex andmultiple partners, but says theyalso have to do more to convince olderwomen to attend. “According to theexperience from other countries, weknow that only the tests that have beengiven during organised screening with aproper follow-up of good quality willdecrease the numbers of women whoare diagnosed with cancer.”

LatviaLatvia had a ‘compulsory’ system ofgynaecological examination in the1980s, but the incidence of cervicalcancer rose once this was abandoned in1989. Cancer rates are especially highin the rural population and a third ofthe women who are diagnosed havestage III or IV disease. A quarter ofwomen die within a year of diagnosis.An opportunistic screening pro-

gramme was launched in 2005, and afull screening programme in 2009, butso far only a quarter of the women whoare invited attend.Only 1 GP in 50 provides gynaeco-

logical care for their patients, and asurvey in 2003 suggested that three-quarters of girls and women aged 15–49 did not trust their GPs to do so.Ilze Viberga, a gynaecologist at Riga

Systems&Services

Women wanted a personal letter inviting them to

screening and to be able to phone for an appointment

Three-quarters of girls and women aged 15–49 did

not trust their GPs to provide gynaecological care

ATTENDANCE IS STILL A PROBLEM

EUROCHIP’s Estonian participantshave been exploring whyattendance rates for cervicalcancer screening remain at lowlevels despite the number ofinvitations going upSource: Department of

Epidemiology and Biostatistics,

National Institute for Health

Development, Tallinn, Estonia

StradinsUniversity, says that this has tochange if the Latvian programme is tosucceed, and she is currently con-ducting a study of doctors’ knowledgeand attitudes.“The general practitioner is not

very interested in this screening pro-gramme, because they think it is thejob of gynaecologists, and the gynae-cologists expect more from the generalpractitioners,” she says, adding, “Wehave to change this philosophy so thatwomen can go to a general practi-tioner, because it does not matter whois going to take this test. Taking asmear does not need specialist skills; itis just simple training. If the result isnot good, then the gynaecologist has tobe involved in the treatment process.”

LithuaniaIn Lithuania, where screening startedin 2004, the response has been a littlebetter, but still less than half of women(44%) attended the first round ofscreening, with the lowest returns inrural areas. Ruta Kurtinaitiene, agynaecologist at Vilnius University, saysthere is a need for a centralised systemof call and recall, with a personalisedletter to every woman.“I think we have a problem with

lack of knowledge and a psychologicalbarrier. I think a woman is scared tocome to a gynaecologist. She does notunderstand that you need a Pap smearevery three years even if you do nothave any disorders or problems withgynaecology. Our early study showsthat if you send a private letter to thewoman, the attendance rates double.”Research conducted by her col-

league Jolita Rimiene for her doctoraldissertation indicated a need for bettertraining in how to do a Pap smear. Shefound that 5%–12% of Pap smearswere evaluated as ‘inappropriate con-tent’ or ‘inadequate’ for cytological eval-uation, and that up to half the cells

collected from the patient never getonto the slide and are discarded withthe test instrument.

BulgariaWith a population of 7.6 million peo-ple, Bulgaria is as big as Estonia, Latviaand Lithuania put together, and with alarger population, screening becomesevenmore of a challenge. The old Bul-garia had a strong tradition of prophy-lactic health checks, but no organisedscreening programme, and when thecountry began to suffer economic hard-ship in the 1980s and 1990s, even thisfell apart. Until the late 1980s cervicalcancermortality rates were comparablewithmany EU countries, but incidencedoubled between 1984 and 2004 forwomen aged 30–49, and the mortalityrates rose 2.5 times.In May 2009, a national Campaign

for the Early Detection of Cancer was

approved, underwhich amillionwomenwould be reachedwith information and200,000 women tested throughout thecountry in 2012. However, the eco-nomic crisis has stalled moves towardsa truly national programme.Yulia Panayotova, from theBulgarian

Health Psychology Research Centre,says there is still a lack of political com-mitment, but she is optimistic that anational programmemaybegin in two tothree years. This early detection pro-gramme includes ‘STOP and GO for aCheck-up’, designed to improve infra-structure, increase capacity and preparesociety for population-based screeningprogrammes for cervical, breast and col-orectal cancers. Improving capacitywillinclude establishing a screening registryand a call-recall system.Panayotova still feels some frustra-

tion at the delays. “Every day a womanis dying from cervical cancer in ourcountry. It is obvious that the best wayis to have an organised programme.There are many people who are takingit seriously but unfortunately we stilldon’t have a programme, whichmeansthat the policy makers are not taking itseriously enough.”

RomaniaRomania has the highest death ratefrom cervical cancer in the whole ofEurope. In 2006, the crude mortalityrate was 20.9 per 100,000 women.Florian Nicula, Head of Epidemi-

ology at the I Chiricuta OncologicalInstitute in Cluj-Napoca, received thePearl ofWisdom award, along with hiscolleagues, for a regional pilot screen-ing programme in Transylvania. Thissaw screening coverage increase fromless than 1% to 20% inCluj county, andsimilar improvements in another fivedistricts.This programme has produced a

‘proven in Romania’model that can beintroduced into the rest of the country.

Systems&Services

Quality control. Jolita Rimiene demonstrated theneed for better training in Lithuania, as up to 12%of Pap smears were too poor to be evaluated

Systems&Services

In particular, it has demonstrated thatwomen in rural areas – always themostdifficult to reach – do indeed wanthigh-quality screening.Nicula recalls how he sent out a

mobile team to a village in a local ruralarea where staff visited women com-munity leaders, convincing them toback the screening initiative. Thesewomen agreed to encourage thewomenin the community to attend a mobilescreening unit, and took a lead by beingthe first to attend. Later in the day hetook a phone call from his team to saythat they would have to stay overnightin the village, because the queues wereso long outside the van.Even so, the 20% success rate is

still far too low says Nicula. “Thewomen responded very well, but wecouldn’t invite all the women weshould, because of logistical and finan-cial problems. The national programmewas supposed to start last year as a rollout of the regional pilot programme, butbecause of the resources crisis we hadto delay the start.”A report on Romania produced for

the EuropeanCervical CancerAssoci-ation points to a lack of political will.“The project hasmade good progress inraising the political priority of cervicalcancer prevention in Romania…How-ever, the majority of politicians still donot understand the complexity of theprogrammes required to achieve goodresults nor the resources that must becommitted to the implementation ofthese programmes.”Building on this regional success,

Romania has appointed the Cluj teamas the National Management Unit

with responsibility to coordinate 21county units across the country andresponsible for quality control for theentire programme.Daniela Coza, epidemiologist at the

I Chiricuta Oncological Institute, saysthat this needs to be a national priority.“It is a huge problem for Romania, aswe have the highest number of newcases and mortality in Europe and oneof the highest in the world. It affectswomen of active ages and women inmiddle- and lower-income groups. Wehave been struggling with this matterfor a long time. Romania has to dosomething for themost at-risk women.”

PolandTaking action alongside EUROCHIP isPoland, which with its population of38.1 million aims to screen 3 millionwomen a year. However, even afterthree years, theNational Cervical Can-cer Screening Programme attracts onlya quarter of thewomenwho are invited.Arkadiusz Chil, from the Kielce

Oncology Centre, says, “Every year,cervical cancer is diagnosed in 4000women in Poland and half of them diebecause of it. These numbers speakfor themselves, which is why we set upour cervical cancer programme. Thereal problem is that cervical cancer isdiagnosed too late in advanced stages.”Magdalena Bielska-Lasota from the

IndependentUnit of Oncological Edu-cation, at theMaria Skłodowska-CurieInstitute of Oncology in Poland, saysthat Polish women lack confidence inthe system. “The programme is organ-ised very well from an administrativepoint of view and it is supposed towork,

but the failure of the screening is thatwe have a very low attendance.“There are a few reasons inmy opin-

ion. One is that there is a crisis withtrust in the system and trusting thedoctors. Women are scared becausethey may have an examination whichdoes not have good quality assuranceand may give false-negative or false-positive results.Mymessage to womenis to press our government and the doc-tors to keep the quality to the levels setby the European guidelines.”There has been a big effort to train

doctors, midwives and nurses. By 2010,7900 professionals had been trained, aswell as 1284 ‘opinion leaders’who, it ishoped, will convince women to attend.Lack of faith in the system is not just

a problem for Poland, but a commontheme in these countries. EUROCHIP,in collaboration with the EuropeanSchool of Oncology, organised mediatraining for key staff in each country tohelp specialists becomemore comfort-able in developing and delivering keymessages through the media.

HPV VACCINATIONThe complicating factors for countriestrying to set up screening services noware the HPV vaccines which hold outsuch promise for the next generation,but also have the potential to demo-bilise efforts for improving existingscreening services. The cost of the vac-cines in the first few years of their avail-ability has also been a constraint in thenew member states, where resourcesare particularly stretched.If given to girls before sexual activ-

ity begins, they have the potential to

His team rang to say that they would have to stay

overnight in the village, because the queues were so long

dramatically reduce the inci-dence ofHPV and thereforecervical cancer.One problem is that it is

unlikely that the benefitswill start to be felt for 15–20years, and the full popula-tion-wide impact wouldtake 50 years or more.Implementing an HPV vac-cination programme is nosubstitute for organising aneffective screening system. The vac-cine is of little value to the populationof women who have already becomesexually active, as it cannot eradicatethe virus where it is already present, orstem the growth of an incipient cancer.And despite very impressive results inclinical trials, they have not yet proventhemselves in country programmes.Romania decided to provide the

vaccines free for girls aged 11 yearsold, and started a school-based cam-paign of vaccination. But the EuropeanCervical Cancer Association reportedthat the take upwas as low as 4% –wellbelow even the worst screening pro-gramme results.Florian Nicula accepts that the vac-

cinecould in theoryprevent almost all thecervical cancers if it reachedenoughgirlsat the right age. In practice, however, in

Romania they have been leftwith stock-piles of the vaccine, which they are nowtrying to use through a new informationcampaign. “The parents did not agree totheir daughters having the vaccine,” hesaid.There is clearly aneed to investigatethepublichealth aspects of theHPVvac-cines; including which implementationmodelswouldbebest for a high coverageand acceptance.ArkadiuszChil fromPoland sees the

vaccine as a distraction from the mainchallenge. “The vaccine is not an alter-native to cytology.Wecannot fight cancerwithout regular cytological examinationand that must be clearly stated.”

IT TAKES TIME TO GET IT RIGHTThere are no short cuts to establishingeffective screeningprogrammes, saysAhtiAnttila –careful planning andattention to

detail are important at every stage.“Even in small country, it takes about

two years to plan everything, taking intoaccount the current activities in health-care and how the screening could beintegrated. One cannot go directly tofull national screening in a short timebecause it is so complicated to get all theparts of the chain into the right order.”The bigger the country, the greater

the challenge. Romania, has a targetpopulation of 6 million women, andeven if theywere only invited for screen-ing every five years, thatmeans 1.2mil-lion invitations a year, probably resultingin a million screening tests and 30,000colposcopies [an investigative diagnosticprocedure usually performed whereabnormalities have been revealed by thesmear test]. Staff have to be trained atevery step of the way, expanding from

Systems&Services

“My message to women is to press our government and

the doctors to keep to the European quality guidelines”

Going for a check-up. This mobileunit covers the villages of

Swietokrzyskie province in Poland,providing both breast and cervicalcancer screening. The system iswell organised but more workneeds to be done on building

awareness and trust, in order toimprove take-up rates

pilot areas, so that they in turn becomereference centres to coordinate trainingand organisational activity in otherregions. “For small countries likeEstonia,Latvia andLithuania one coulddo every-thing in 5–10 years. For a large countrylikeRomania it could be evenmore than10 years.”Anttila points out that even in Fin-

land they still do monitoring and eval-uation. “The Finnish programmestarted in 1962 and has continued foralmost 50 years, but every year we stillcollect data. The cancer burden isextremely low when screening workswell, but this systematic monitoring

and learning has to be part of it.” TheFinns are also proactively studyingpotential newmethods for cervical can-cer prevention, such asHPV screeningor HPV vaccinations.With the EUROCHIP work pack-

age due to conclude at the end of 2011,the organisers are to ask the EuropeanCommission for more time to addressscreening problems, which would allowtime for Bulgaria and Romania to fur-ther develop their pilot schemes.AndreaMicheli, leader of EUROCHIPfrom the Italian Istituto NazionaleTumori in Milan, believes the workconstitutes a vital step towards reduc-

ing cancer inequalities across Europeand so reducing the overall burden ofcancer.Anttila says they should not step

back now. “We do not yet have screen-ing programmes of high quality. Wecannot say we have them, and it hasnot been adequately resolved. I don’tthink the EU would want to give up.“All these countries have made a

start and need to consider whatmore todo. Then, political commitment willcome, aswe’ve seen happen in countrieslike England, and everything will workvery effectively. Maybe it takes evenmore than 10 years, but I think theinformation that EUROCHIP hasshared in the countries and scientificcommunities takes us one step further.”Screening is a major focus of the

recently launched European Partner-ship forActionAgainstCancer, he adds,which reinforces the political will tomake progress on this front at the high-est political level of the Union.Micheli also believes that the data

and experiences being provided by thesecountries is like gold dust. “Themost pre-cious element we had in EUROCHIPwas the availability of data to allow com-parisons amongst all the countries ofEurope. Through solidarity and net-working, countries are now sharing expe-riences on best practice and developingscreening programmes at lower cost thanif they were tackling this alone.”

Systems&Services

“It takes time because it is so complicated to get all

the parts of the chain into the right order”

Cancer World 42

Documents

Transcript of Cancer World 42

Cancer World 65

CANCER FACT SHEETS: PROSTATE CANCERgco.iarc.fr/today/data/pdf/fact-sheets/cancers/cancer...CANCER FACT SHEETS: PROSTATE CANCER Estimated age-standardised rates (World) in the world

Cancer World Issue 55

World Without Cancer

A World Without Cancer

Cancer World 36

Cancer World 47

Cancer Prevention Month and World Cancer Day Social Media ... · Cancer Prevention Month and World Cancer Day Social Media Toolkit . is designed to help public health professionals

Cancer World survey on Cancer Care during the Pandemic ...

World Cancer Day 2017 – Fight Against Cancer And Save Lives

Gastric Cancer in the World

Cancer World 41

INTERNATIONAL BREAST CANCER STUDY GROUP IBCSG 42 …

Cervical Cancer Treatment at World Best Cancer Hospitals in India

world cancer research fund

Cancer World 43

FESPA WORLD Issue 42 - Español

The cancer process - World Cancer Research Fund International · The Continuous Update Project (CUP) is World Cancer Research Fund Network’s ongoing programme to analyse cancer

World Outlook Issue 42

Cancer World 50